Chromosome Abnormalities and Genetic Counseling , 3rd Edition

24.Preimplantation Genetic Diagnosis

Chromosomal preimplantation genetic diagnosis (PGD) is done in the setting of in vitro fertilization (IVF), and in principle enables an unaffected embryo to be transferred to the uterus, some 3 or 4 days post-fertilization. Thus, for couples facing a high genetic risk, the risk can be bypassed, and the prospect of pregnancy termination for the reason of genetic abnormality can be avoided. Five major categories (some of which may overlap) of patients coming forward for consideration of chromosomal preimplantation genetic diagnosis are the following:

· One member of the couple having a balanced chromosomal rearrangement

· Some women having had multiple miscarriages

· Subfertile women of older childbearing age

· Some male infertility for which intracytoplas mic sperm injection (ICSI) is used

· Failed previous attempts at embryo transfer following IVF

Developments over the past decade with IVF, human embryo culture and manipulation, molecular genetics, and FISH are the factors that have made PGD possible. From an essentially research-based exercise in a very few laboratories in the early 1990s, it has advanced to being, in the early 2000s, a diagnostic tool available through a number of larger IVF clinics. Chromosomal PGD is typically done on a blastomere (single cell) from a 6–10 cell embryo¹ at day 3. The selected embryo is then transferred, on the same or next day. In principle, provided the diagnosis was accurate, and barring the presence of mosaicism (see below), the pregnancy can proceed in the knowledge that the baby will be unaffected. There are two main categories of chromosomal PGD: PGD being done as a general aneuploidy screen (PGD-AS), and focused PGD with respect to a particular parental translocation, or other rearrangement.

While the general principles of counseling for chromosome abnormalities, as outlined elsewhere in this book, are relevant to couples considering PGD, there are some special considerations that require elaboration. Chromosomal PGD is technically challenging for the scientists who do the work and emotionally demanding for couples who want to benefit from it (and not inexpensive). The issues are dealt with in detail in the book Preimplantation Genetic Diagnosis (Harper et al., 2001), to which the reader wishing fuller information is referred, the chapter on chromosome abnormalities by Delhanty and Conn being particularly germane. Some of the practicalities of the logistics of PGD are also discussed in the paper of Geraedts et al. (2001), “a collaborative activity of clinical genetic departments and IVF centers” coming from twelve countries. Reviews are provided by Egozcue et al. (2002), Kanavakis and Traeger-Synodinos (2002), Munné and Wells (2002), and Wilton (2002), and detailed analyses of the accumulated experience of a number of clinics are documented in the annual reports of the European Society for Human Reproduction and Embryology (ESHRE) PGD consortium steering committee, which appear in Human Reproduction. Writing in 2002, Lavery et al. reflected as follows: “PGD is not an easy solution for high-risk genetic couples. This technique is still in its infancy and it should be remembered that both chorionic villus sampling and amniocentesis were performed for many years before they became clinically accepted. Our data suggest that PGD is acceptable to patients and is a valuable alternative to prenatal diagnosis.” This cautious optimism is countered by a more restrained view in Egozcue et al. (2002), who focus upon the statistics of actual outcomes, which, it is true, fall well short of what one might wish. Those embarking upon PGD do need to have a realistic appreciation of what the odds for success might be, according to their particular clinical circumstances.

Concerning the risk percentages noted in this chapter, it is to be borne in mind that the assessment of aneuploidy in the early embryo is usually based on analysis of a limited number of chromosomes, and so some of the figures are likely to be lower estimates. Most FISH is designed to probe five chromosomes, nos. 13, 16, 18, 21, 22; or, if two rounds of FISH are used, bringing in chromosomes 14, 15, X, and Y, nine chromosomes are testable—“5-color FISH” and “9-color FISH,” respectively.

PATIENTS FOR WHOM CHROMOSOMAL PREIMPLANTATION GENETIC DIAGNOSIS MAY BE APPROPRIATE

Carriers of Balanced Rearrangements

A parent who is the carrier of a balanced rearrangement typically has a high risk of producing unbalanced embryos, as discussed at length in earlier chapters. Particularly in the context of an unfortunate reproductive history, often with several miscarriages, or with one or more terminations following conventional prenatal diagnosis of an unbalanced fetal karyotype, the attraction of PGD is obvious. Where, as is often the case, there is infertility which necessitates IVF and possibly ICSI, the embryo is of course already accessible, and PGD is applied as an additional procedure. The two main categories are reciprocal (rcp) and Robertsonian (rob) translocations. Munné et al. (1998c) demonstrated that translocation carriers having had PGD were much less likely to have a spontaneous abortion. In their series of 35 cases, the abortion rate fell from 92% without PGD to 13% following PGD—a rather impressive statistic. This considerable improvement does imply that many of these couples would otherwise have had no impediment to fertility. Pregnancies have been successfully achieved, although they are well outnumbered by failed attempts, as applies, of course, to all IVF. In the ESHRE (2002) report, on the basis of data from procedures undertaken in a number of clinics worldwide during 2001, the pregnancy rates per embryo transfer (pregnancy according to detection of heartbeat) were 23% (rcp) and 29% (rob).

Women of Older Childbearing Age

It would scarcely be practicable to offer PGD as a standard procedure to mothers entering the age group in which the risk for aneuploidy is becoming more substantial. But there may be some for whom PGD-AS will be reasonable, more often in the setting of difficulty in conceiving, as ovarian function is beginning to wane, or with a history of miscarriage. In other words, older mothers who might otherwise have considered IVF would have the option of adding PGD-AS to the procedure, to improve their chances of having a normal child. The success rate (pregnancy per transfer) in these women is 36%. This figure, a good one by IVF standards, is from the ESHRE (2002) data for 2001. The improvement is greater in women over age 39 years (Table 24-1).²

Recurrent Miscarriage

Couples who are chromosomally normal but who have had a history of multiple miscarriage may benefit from PGD-AS (Munné et al., 2000a; Rubio et al., 2003). The fraction of spontaneous abortion that is due to chromosomal abnormality has been revised upwards in recent years (Fritz et al., 2001b; and see p. 344), and some miscarrying couples may have been having recurring aneuploidies, whether from meiotic or very early postzygotic mitotic error. For such couples, the success rate, according to the ESHRE (2002) analysis, is 32%. Of the embryos biopsied in this series, 56% were abnormal (not an unusually high fraction in this setting).

Infertility

Around half of all embryos from couples undergoing IVF for infertility are chromosomally abnormal. The proportion of morphologically normal embryos with chromosome abnormality is about 20%, and in arrested embryos, as high as 70% (Hardy et al., 2002). Use of PGDAS allows the couple to avoid transferring an embryo with a coincidental aneuploidy. In a comparison of women 35 years and older who were undergoing IVF, some of whom had PGD-AS and some did not, the former had fewer spontaneous abortions and more liveborn children than did the latter (Munné et al., 1999). Of course this outcome is not surprising, with the presumed aneuploid embryos having been discarded. The issue of the chromosomal prognosis in pregnancies from an ICSI conception, the man having oligospermia, remains to be clarified (Silber et al., 2003). A translocation might be the reason for the infertility, and this possibility should be checked ahead of proceeding with ICSI (Veld et al., 1997).

Previous Implantation Failure at in Vitro Fertilization

The overall failure rate at routine IVF is considerable, with only about a quarter of embryo transfers, on average, leading to a clinical pregnancy. Among these patients, there are women who have 0% success, in spite of having had many attempts. In some of these women, maternal factors will predominate (“uterine receptivity”), but in others the reason may be related to a continuing production of chromosomally abnormal embryos. Gianaroli et al. (1997) studied 36 such patients with repeated IVF failure or other reason for poor prognosis to achieve a pregnancy, and made the observation that about 60% of embryos were karyotypically abnormal. While this figure is fairly unremarkable in the context of IVF, more notably, mosaicism and polyploidy were the predominant abnormalities in these embryos, rather than simple aneuploidy. The more failed attempts there had been, the higher was the aberration rate.

The choosing of a normal embryo would seem a logical solution for those women suffering failed embryo transfer, but in fact the odds may not be much improved with PGDAS. Overall in this group, only 11% of embryo transfers resulted in clinical pregnancy in the ESHRE (2002) report. This figure compares with the rates of 36% and 32% in the groups of mothers over age 35, and of couples with recurrent miscarriage, respectively, as noted above. Some groups, however, do report better success with the use of PGD-AS in “poor prognosis” patients (Wilton, 2002), possibly reflecting selection criteria.

CLINICAL AND LABORATORY PROCEDURES

From the above categories of patient, two karyotypic classes of couple are to be considered: those who have normal chromosomes, and couples in whom one member carries a balanced chromosomal rearrangement. The latter group will command most of our attention, although they are, of course, the smaller group. Those who make the decision to embark upon chromosomal PGD, and for whom the laboratory staff have advised that testing, in their particular case, would be feasible, will need to enroll (if not already) in an IVF program. Close liaison with the laboratory is required, so that the scientist can have the appropriate FISH probes prepared and ready for use on the day. The time frame for having hyperstimulation of ovulation is established from the woman's menstrual cycle, and ovum “pick-up” is conducted by transvaginal endoscopy under ultrasound guidance. Given the high risk of abnormal malsegregants typically associated with chromosome rearrangements, it is useful to employ a stimulation protocol designed to maximize the numbers of ova produced (Fridström et al., 2001). The ova are collected and then exposed to sperm in vitro. If the couple are of otherwise normal fertility, simple mixing with the male partner's sperm suffices. With some forms of male infertility, and this is quite often the case in the male heterozygous for a chromosome rearrangement, ICSI is needed.

On day 1, around 18 hours after exposure to sperm, the oocytes are checked for the presence of two pronuclei and two polar bodies, as evidence that fertilization³ has occurred. They are then returned to tissue culture medium; in a few hours syngamy will occur, and over the next 48 hours the first few mitoses will have produced cleavage-stage embryos of 6–10 cells. On day 3, in the morning, one or at most two cells (blastomeres) are removed from each embryo, under the dissecting/manipulating microscope.⁴ This requires a hole to be made in the “shell” (the zona pellucida, which has not yet been cast off), the cells being extracted by very gentle suction. These cells are subject to FISH (or possibly other) analysis to determine whether they have a balanced or unbalanced form of the rearrangement. One or two embryos tested to be chromosomally normal (or balanced) are transferred to the uterus on the afternoon of day 3 or the morning of day 4 and, with luck, one⁵ will develop into a normal infant. The remaining embryos with a normal/balanced chromosomal complement will probably be cryopreserved, in case the first embryo transfer does not result in a pregnancy, and perhaps for a second pregnancy further in the future (although the pregnancy success rate post-thawing is less). The process is outlined in Figure 24-1.

If the embryo is incubated for 1 or 2 more days, it advances through the morula and early blastocyst stages. Considerable selection pressure applies during this short period, and many chromosomal abnormalities, including most monosomies and extensive mosaics, impose a lethal burden (Clouston et al., 2002). The chances of successful transfer may be better if the embryo has declared itself capable of developing this far (Tao et al., 2002; Van Der Auwera et al., 2002). In principle, there might seem some attraction in delaying PGD until the blastocyst is forming, since cell number has increased and differentiation between inner cell mass and trophoblast has begun, allowing sampling to be focused on the trophectoderm (in a sense, a very early CVS) (De Vos and Van Steirteghem, 2001). This would involve making a hole in the zona pellucida and allowing a small part of the lining of the blastocele cavity to herniate through, and this tiny bulge could be excised by laser. The technique is presently in a developmental stage (De Boer et al., 2002). A more controversial question is whether longer incubation in vitro may disturb the epigenetic state of the embryo.

Initially, PGD relied primarily on the use of DNA amplification methodology (PCR) either for the assay of embryonic sex or to detect single gene mutations. Preimplantation genetic diagnosis for chromosome abnormalities, as developed over the past decade, is based primarily on FISH technology. Recently, comparative genomic hybridization (CGH) has been added to the diagnostic armamentarium (Wilton et al., 2001) (see below).

The polar body (PB), as a mirror image surrogate of the ovum, has been used in some research PGD laboratories (Munné et al., 1998d; Verlinsky and Kuliev, 2000). By way of example, imagine that the asterisked gametocyte in Figure 2-3a is the first PB (PB1), and that the two chromosomes shown within it are no. 18s—that is, PB1 is disomic 18. The “empty” gamete to the right, therefore, would be a nullisomic 18 oocyte, and thus, of course, be discarded. A nullisomic second PB (PB2) (one of the empty cells in the next row) should provide corroboration. The reader may also determine, on study of Figure 2-6 with respect to predivision of chromosomes at meiosis, why analysis of PB2 alone could in some instances be misleading. The cell labeled “disomic gamete” in this figure could be the oocyte, but PB2, represented by the cell next to it, shows a normal monosomy. Both PBs together can enable the full picture to be deduced, and the disposition of all four chromatids can be accounted for, provided probes are used that enable the distinction between single chromatids and double-chromatid chromosomes. Removal of the PB appears to have no effect on the child subsequently born (Strom et al., 2000). However, while PB diagnosis is an elegant approach, satisfyingly requiring recall of some elementary facts of biology, most PGD laboratories are likely to remain with the blastomere. Furthermore, PB-based PGD assumes the sperm to have a normal haploid chromosome constitution, but this would be untested. Finally, some chromosomal errors occur after the events of meiosis, at an early postzygotic stage.

SEGREGATION ANALYSIS AND PROBE SELECTION

Analysis of segregation patterns of structural rearrangements has been discussed extensively in earlier chapters. The specific PGD-related risks for reciprocal and Robertsonian translocations are noted in Chapter 4 (p. 65) and Chapter 6 (p. 126), these two forms accounting for the substantial majority of “chromosomal” PGD patients. When choosing FISH probes for a particular PGD, each and every possible segregation outcome, as noted below, must be considered. The pattern of FISH signals that each outcome would generate and the certainty of being able to distinguish a balanced or normal chromosome constitution need to be carefully thought through. For the common case of the autosomal translocation, and using the example of a translocation 46,XY,t(14;18), we set out in Color Figure 24-2 (see separate color insert) the full range of possible blastomere combinations. Pericentric inversions lead only to two unbalanced forms (p. 145), and these can be accounted for by the use of a subtelomeric probe at either end of the chromosome, and a centromeric probe (Escudero et al., 2001). As noted above, the more embryos that can be biopsied, the better, since the odds for unbalanced embryos are high, and may vary unpredictably from cycle to cycle (Fridström et al., 2001).

Probes are designed which hybridize to judiciously chosen parts of the chromosomes of interest. With most simple reciprocal translocations three probes will, in general, be required: two that hybridize to a point within the translocated segments, and one to one of the centromeres. The most readily available commercial markers for recognition of the translocated segment are subtelomeric probes (Scriven et al., 1998). Observing the number and (to some extent) the disposition of colored spots in the nucleus of a blastomere removed from the IVF embryo allows a deduction of the chromosome complement. Probes to α-satellites enable rapid (2- to 3-hour) detection of the centromeres, although the single- or low-copy probes for telomeres take longer to hybridize, necessitating a late day-3 or early day-4 transfer (Delhanty and Conn, 2001). It is necessary to show that the probes give clear and unequivocal signals on interphase nuclei and chromosomes from the carrier parent, ahead of proceeding with IVF and PGD. If there is a living individual with the imbalanced state, the opportunity should be taken to check that the probes would give a correct interpretation. These points on probe selection are dealt with more expansively by Delhanty and Conn (2001).

The Technical Challenge of Fluorescence in Situ Hybridization

Single (or at most two) cell FISH requires consummate skill on the part of the cytogeneticists and embryologists who do this work. Even the most technically adept scientist, however, cannot achieve 100% of resolvable FISH signal on all chromosomal targets. This is not normally a problem for other applications where many nuclei or metaphases are available for study. In PGD it is an important limitation.

Consider the impediments to success in a “simple” PGD for a typical reciprocal translocation. Suppose three probes are to be used, each probe hybridizing to two chromosomal sites (on two normal chromosomes, or on the normal and derivative chromosomes, respectively), to give six hybridization spots in total from the normal or balanced embryo. Supposing that the hybridization efficiency of each probe is 98%, the probability of the normal or balanced embryo giving six correct signals in one cell is (0.98)⁶ = 0.89. These facts of simple arithmetic point to the risk that a normal or balanced embryo could be diagnosed as unbalanced, and vice versa, purely because hybridization is less than 100% effective. Munné et al. (2000b) estimated an average error rate of 6%, based on studies of all blastomeres in a series of donated embryos, and 10% in real PGD situations. Signal splitting, signal overlap, and incompletely penetrating probe may be the usual reasons. The use of two cells would reduce the risk of errors; if the two differed, the diagnosis would be based on the nucleus that had the most signals. The impact of this on any specific diagnosis would need to be determined on a case-by-case basis, but the most likely adverse outcomes would be wastage of normal embryos (diagnosed as partial monosomies) and diagnosis of embryos with a partial trisomy as normal. It is not surprising that centers that offer PGD usually recommend that it be followed up by conventional prenatal diagnosis to check that these sorts of error have not happened (Munné et al., 2002a).

Comparative Genomic Hybridization

Comparative genomic hybridization (CGH) (p. 12) is an alternative and yet more demanding procedure, although it does have the advantage of not requiring specific or tailor-made probes (Voullaire et al., 2000b). The whole karyotype is in principle examinable. There is the disadvantage that the time taken for analysis means that embryos have to be cryopreserved after biopsy, pending the result. As with FISH, the normal and the balanced carrier states cannot be distinguished. Also, with CGH the resolution of small segment imbalance may not be possible, and thus its suitability for translocation cases is limited. The first infant born following PGD-AS by CGH was reported in 2001 (Wilton et al.). Advances in this methodology may enable its routine use and, potentially, improve the chances of having a healthy child (Wilton, 2002). In a tour de force that is unlikely to become a routine, Wells et al. (2002) describe the CGH approach applied to polar bodies, in a 40-year-old woman who had suffered previous IVF failure. Of 10 analyzable polar bodies, all but one were aneuploid, and the embryos therefore predicted to have the countertype aneuploidy. Transfer of the single normal embryo did not succeed.

CONSIDERATIONS OF EMBRYOLOGY

The Problem of Mosaicism and “Chaotic” Embryos

The 6–10 cell embryo (the stage at which blastomere sampling is typically done) probably contains only one or two cells whose descendants will go on and form the inner cell mass and, thus, eventually the fetus. Chromosome studies on IVF embryos can reveal different chromosome constitutions in different cells, up to the point of “chaotic” embryos in which several cells each have a different aneuploidy (Handyside and Delhanty, 1997; Iwarsson et al., 2000; Sandalinas et al., 2001); it is not necessarily easy to guess, from the observed pattern of the different aneuploidies, what might have been the sequence of events at each individual mitosis that was able to lead to this eventual picture. Munné et al. (2002b) list these four main categories: diploid/polyploid mosaicism, chaotic mosaicism, mosaicism due to mitotic nondisjunction, and “split” mosaics with two cell lines that complement each other. Most mosaics originate in the first three mitotic divisions: across all categories, 50% at the first mitosis, 25% at the second, and 25% at the third. The level of mosaicism may increase going from the cleavage stage through to the morula (Bielanska et al., 2002a).

It may be that a particular vulnerability applies to these very early mitoses, before the necessary genes for cell-cycle checkpoint control have fully swung into action, and maternal cytoplasmic factors are being relied on (Hardy et al., 2002; Voullaire et al., 2002). Alternatively, or perhaps additionally, there may be a male factor involved, with impairment of the embryo's centrosome function. This could apply more particularly to cases of a severe spermatogenic defect, with a poor quality sperm bringing a poor quality centriole to the embryo, given that the first few mitoses make use of the centriole that came with the sperm (Silber et al., 2003). The phenomenon of mosaicism may be patient-specific, reflecting an individual predisposition (Delhanty et al., 1997). Mosaicism is more often observed in embryos of translocation carriers, and possibly the presence of a translocation chromosome may of itself incline toward the generation of mitotic malsegregation (Iwarsson et al., 2000; Emiliani et al., 2003).

Mosaicism is a surprisingly common observation. In a study of 12 “normal” appearing embryos that had been frozen beyond a statutory time of 5 years and thus available for research, and using a CGH approach, Voullaire et al. (2000b) diagnosed only three as having non-mosaic normal karyotypes. Four of the embryos had a mixture of normal and aneuploid cells. The proportion of chromosomally normal embryos found here is lower than that in other studies using chromosome-specific FISH, since CGH detects (in principle) the full spectrum of aneuploidy. Similar results from a CGH study of another 12 embryos were reported by Wells and Delhanty (2000), who found only three to be apparently normal, one to be aneuploid, six to be simple mosaic, and two to be chaotic mosaic. A yet greater abnormal proportion, 100%, was seen in a study of 22 spare embryos post-PGD from seven translocation carriers (Malmgren et al., 2002). The blastomeres of some embryos were balanced for the translocation, others were unbalanced, and each embryo contained some cells with various other abnormalities. Some embryos showed extreme chaotic mosaicism, with a different chromosomal complement in every blastomere. Six of six spare embryos from presumed chromosomally normal parents also showed aneuploidy and/or mosaicism.

This mosaicism cannot necessarily be dismissed as a phenomenon affecting some cells but leaving normal ones to carry on through. Magli et al. (2000) showed that mosaicism on blastomere biopsy could be reflected in mosaicisms, often of evolving complexity, within the inner cell mass itself (and see p. 343). These observations are also made in embryos of the translocation carrier, the majority of which may be mosaic. Nevertheless, it may be that the window of observation afforded while the embryo is in vitro is a somewhat distorting window, in terms of foreseeing the eventual outcome for the fetus. These observations of chaotic mosaicism are made on embryos created artificially, usually from infertile couples using superovulated eggs, the male partner often being oligospermic, and maintained in an artificial environment for the period prior to the chromosome analysis. It is plausible that associated stresses could bear unfavorably on the process of mitosis, compared with embryos generated in vivo—that is to say, conceived naturally.

The Need to Consider a Full Range of Abnormal Segregant Outcomes

Postzygotic selection against aneuploid embryos has had little chance of having operated by day 3. Thus, in the case of translocation carrier parents, all possible segregant outcomes of a rearrangement may be encountered (Fig. 4-4, Table 4-2), and FISH probes will need to be selected accordingly. Some complete or partial autosomal monosomies, practically unknown otherwise, may be associated with occult abortion in the earliest days post-conception (p. 343). Conn et al. (1999) describe PGD in the setting of a parental t(6;21), using a particular FISH strategy that covered the possibilities for chromosome 21 imbalances, and one embryo had a normal signal. This embryo was duly transferred, but with no more than a “biochemical pregnancy” resulting. In fact, this may have been interchange monosomy 6. In chromosomally normal couples of whom the woman is of older childbearing age, more embryos with aneuploidies are likely to be generated. A maternal age effect with respect to trisomies 1, 16, and 17, scarcely a concern in usual maternal age counseling, might have some relevance in the setting of PGD (Bahçe et al., 1999). Even the spectacular circumstance of a 4:0 malsegregant embryo may be encountered.

GENETIC COUNSELING

Many counselors working in the field of infertility have moved to accommodate the new dimension that PGD has brought about, and they have largely been responsible for dealing with the patients for whom the procedure may be helpful. While it has been true that “PGD is more widely known in the reproductive medicine community than in [clinical] genetic circles” (Geraedts et al., 2001), a fruitful coming together of the two disciplines should be to the advantage of both, and to the patients who attend the clinic.

Preimplantation genetic diagnosis is sufficiently complicated, not to mention expensive, that it will not usually be the first option for fertile couples wishing to avoid the birth of a child with a chromosomal disorder. High-risk scenarios, might, however, warrant early consideration. For infertile couples (whether or not there is a chromosomal basis of the infertility) who require an IVF procedure to conceive, PGD-AS offers the hope of fewer spontaneous abortions and thus a higher likelihood of achieving a successful pregnancy in any cycle. A number of points need to be raised with couples who present for PGD one of whom is the carrier of a chromosomal rearrangement.

1. The reasons for choosing PGD as an option. Some couples may have had conventional prenatal diagnosis with successive terminations of pregnancies due to a high-risk translocation, and are unwilling to face this prospect again. It may be difficult to distinguish a run of bad luck, with an optimistic outlook for the next pregnancy still a realistic possibility, and therefore allowing the counselor to suggest a further natural attempt. Or, the series of abnormal pregnancies may reflect a strong predisposition of that translocation to generate unbalanced gametes. Avoiding the possibility of termination following conventional prenatal diagnosis is, for those who have had that experience, a strong motivation (Lavery et al., 2002). Those who have suffered several failures of embryo transfer (a sort of very early miscarriage) comprise another group for whom PGD may have appeal.

2. The limited success rate. As discussed above, many IVF/PGD procedures do not produce the desired end result of a child. Referring to the ESHRE data for 2000–2001, couples presenting for the reason of a parental translocation (rcp or rob) underwent a total of 159 egg collection cycles. Of the 1190 embryos produced and then successfully biopsied, a diagnosis was obtained in most (89%), and 287 (27%) had the potential to be transferred (that is, the karyotype appeared normal or balanced). Of the 123 embryos actually transferred, a successful pregnancy resulted from 34 of these. Thus, based on the number of egg collection cycles, a pregnancy was established in 21%, and based on the number of embryo transfers, the success rate was 28%. The opposite side of the coin is a 79% failure per cycle, or 72% per transfer. These figures are, however, similar to those applying to all IVF patients. Thus couples who would otherwise have no difficulty conceiving should weigh the pros and cons of PGD and conventional prenatal diagnosis (Kanavakis and Traeger Synodinos, 2002).

For many, the process of IVF can be an emotionally draining experience, and the desired aim of a normal child is not necessarily achieved. Only small numbers of ova may be collected and none be suitable for transfer, transfer may not lead to pregnancy, and pregnancies may miscarry. Counselors seeing these couples need to be knowledgeable about all aspects of IVF and PGD process, including an understanding of their local success rates.

3. The specific genetic risk. The figures provided elsewhere in this book largely relate to the risk for an unbalanced chromosome complement in either a liveborn child or at conventional prenatal diagnosis. The risk that an embryo at PGD will be abnormal is a lot higher. In the experience of Munné's (2000b)group, reciprocal translocation carriers had a 79% abnormality rate on embryo chromosome study and a 70% rate with the rob carrier (the latter all male), these figures including mosaics and embryos in which the segregation mode was not determinable.

The figures for the rcp carrier in Mackie Ogilvie et al. (2002) in Table 4-2 show that, of those embryos in which the segregation mode could be deduced, about half were abnormal. It is becoming apparent, and it was certainly expected that this would be so, that the risks may vary considerably according to the specific translocation. The figures just noted may be higher than in the generality of translocation carriers, but may well be appropriate for the subset who present to the IVF clinic with an unfortunate reproductive history. For recurrent rearrangements, such as Robertsonian translocations and the t(11;22)(q23;q11), data may be pooled to give PGD-specific risk figures (Table 24-2). Fridström et al. (2001) note that while Robertsonian carriers may have a better chance than rcp carriers that an embryo post-PGD would be suitable for transfer (about 30% cf. 10% in their clinic), the pregnancy rates per transfer are similar for the two groups (20%–25%).

4. Follow-up in the pregnancy. Understandably, some couples will be unenthusiastic about an invasive procedure that could possibly put at risk the pregnancy in which there has been so much investment (Meschede et al., 1998b). Nevertheless, couples need to be aware that chromosomal PGD cannot provide a guarantee, both in terms of the potential for diagnostic error (Munné et al., 2002a, report a 2% false normal diagnosis rate) and for there having been mosaicism of the embryo. Prenatal diagnosis should be offered. Ultrasonography may be an acceptable, if imperfect compromise, only proceeding to CVS or amniocentesis if risk factors are detected. Maternal serum screening offers a further possibility (Ludwig et al., 2001), although Maymon and Jauniaux (2002) report slight differences in PAPP-A levels in comparison to naturally conceived pregnancies, and they note also the higher frequency of multiple pregnancy from IVF, which would also complicate the interpretation.

5. Nature may intervene. A natural pregnancy may be achieved while the couple waits for the IVF/PGD preparations to be made. For example, the adjacent-2 karyotype shown in Figure 4-10 came from culture of the products of conception of this couple's third miscarriage, and no normal pregnancies, the woman being a t(13;16) carrier. The outlook did not seem very promising, and plans were being put in place for IVF, but they then reported a naturally conceived pregnancy, in which amniocentesis showed a 46,XY karyotype.

Notes

1. We mentioned the power of language in referring to “babies” or “fetuses,” and “mothers” and “pregnant women,” on p. 374. Embryo is another word with a laden identity, and expressions such as “embryo destruction” can carry emotional weight. In fact, the embryo proper has yet to develop—this only happens when the primitive streak has started to differentiate. Jones and Veeck (2002) debate these fine points, and propose that pre-embryo is the more accurate term for the period from conception through to the appearance of the primitive streak, that is, from 0 to 14 days. Another word with credible currency for about day 4 is morula. On day 5, the next stage is the blastocyst (and see Color Fig. 25-1, 1 and 2).

2. Couples choosing oocyte donation, and if the donor is in her mid 30s or older, may also wish take advantage of PGD-AS.

3. Since fertilization in vitro can be observed as it actually happens, the fine detail of the process can be appreciated. The first act is penetration of the ovum by the sperm. To the embryologist, this is only the prelude to conception; the true moment of conception is the point at which the male and female pronuclei fuse, their chromosomes aligning on a common metaphase plate (syngamy). Once that event has taken place, the zygote has come into existence. At the first mitosis, it loses that name, and becomes, in IVF parlance, a “cleavage stage embryo,” or simply an embryo (or more pedantically but perhaps usefully, as commented above, a pre-embryo).

4. Removing two cells from an embryo with seven or more cells apparently does not affect the potential of the embryo to develop (Van de Velde et al., 2000).

5. If two embryos are transferred, this is not designed to produce twins, but rather to improve the odds that one will succeed. Preimplantation genetic diagnosis may allow a lesser number of embryos to be transferred, ideally just one, thus reducing the likelihood of multiple pregnancy.