REVIEW: Pre-implantation genetic testing



Since its first clinical use, pre-implantation genetic testing has become increasingly available as a method of detecting genetic disorders before pregnancy is established. Based on a difference of purpose, pre-implantation genetic testing can be classified as either pre-implantation genetic diagnosis (PGD) or pre-implantation genetic screening (PGS). Pre-implantation diagnosis was established in the early 1990s as an alternative to prenatal diagnosis and termination of pregnancy.1 Pre-implantation genetic diagnosis describes the testing of oocytes or embryos from patients who have a significant risk of conceiving a pregnancy affected by a known recurrent genetic disorder, in order to pre-select for transfer only those embryos found not to carry the disorder. Pre-implantation genetic screening, on the other hand, aims to improve the outcome of assisted reproduction treatment for the subfertile, by testing for a number of the more frequent chromosome aneuploidies in an attempt to improve implantation and reduce the incidence of miscarriage. It is estimated that more than 1000 children have been born following pre-implantation genetic testing.2 The clinical application of pre-implantation genetic testing is expanding into new areas creating novel ethical and practical dilemmas.

In order for a pre-implantation genetic test to be performed, a representative sample of the embryo is required. A single cell can be removed from a late cleavage stage embryo (8–16 cells), or a few cells may be taken from the trophectoderm of a blastocyst, or one or both of the polar bodies can be removed from the unfertilised egg or early zygote. First polar body biopsy has advantages in that the test is effectively done pre-fertilisation, and samples extra-embryonic tissue and thus may be less likely to affect cleavage of the embryo.3 However, only information about the maternal genotype can be gathered, and since the diagnosis in the oocyte is inferred, crossover during meiosis may render the result less reliable. For these reasons, a complementary cleavage stage biopsy may be required. Cleavage stage biopsy is the most commonly used technique where embryos are grown until they reach the six- to eight-cell stage on day three after insemination and one or two blastomeres removed (Fig. 1). Blastocyst biopsy overcomes some of technical difficulties of molecular diagnosis as more cells are available to be removed from extra-embryonic tissues, but culturing embryos to the blastocyst stage may reduce the cohort for biopsy as only 36% of embryos mature this far, and the time available for diagnosis is limited by the need for transfer on the same day as biopsy.4

Figure 1.

Cleavage stage biopsy. This shows a cleavage stage embryo 72 hours post-ferilization held stationary on a glass micropipette (left) by gentle suction. The zona pellucida has been breached by acid Tyrode solution. A single nucleated blastomere has been removed using a suction pipette (right). Note the clear single nucleus in the blastomere.

Pre-implantation genetic diagnosis

The first successful clinical application of pre-implantation diagnosis to avoid the X-linked condition adrenoleucodystrophy, used PCR to amplify a specific repeat on the Y chromosome in order to sex embryos.5 Proof of principle had been demonstrated in 1968 when rabbit embryos were sexed using sex-specific chromatin following blastocyst biopsies.6

Before the advent of pre-implantation diagnosis there were few reproductive options for couples at high risk of transmitting a serious genetic disorder. These included accepting the risks of the condition and hoping for favourable odds; opting for prenatal diagnosis (amniocentesis or chorionic villous sampling) and then being prepared to terminate an affected pregnancy—an option not taken lightly as it is often performed in the second trimester with considerable physical and psychological morbidity; or the use of gametes from donors who are not carriers of the disorder. In addition, for X-linked conditions, there is now the possibility of using fluorescence activated cell sorting, to sort semen to separate X- and Y-bearing spermatozoa. Although this process is not totally reliable, it may be used to skew the result in favour of an unaffected pregnancy.7 Couples may otherwise chose adoption or decide to remain childless.

Patients choose pre-implantation diagnosis for a variety of reasons. They may already have an affected child, or have lost affected children due to the genetic condition and wish to avoid further risk. They may have had several terminations of affected pregnancies and wish for another method to eliminate or minimise that outcome, or they may have personal, including religious objections to termination of pregnancy. Some may carry a dominant disorder and as such, either are, or will become affected by that condition, and wish to avoid having a child affected as they are, and also wish to eliminate the condition from their ‘bloodline’. They may have seen the genetic condition affect their family and wish to avoid the risk before attempting pregnancy themselves. They may have suffered multiple miscarriages due to a chromosome rearrangement and see pre-implantation diagnosis as an option to reduce the risk of further miscarriage.

Clinical considerations

In order to offer a safe and effective service, a multidisciplinary team should be established. This should include genetic and infertility counsellors, clinical geneticists, cytogeneticists, molecular biologists, embryologists and reproductive specialists. Those performing embryo biopsy must be approved by the Human Fertilisation and Embryology Authority (HFEA) after inspection. Only 12 clinics are currently licensed for biopsy in the UK and fewer for full processing of biopsy sample for diagnosis. Laboratories should participate in external quality assessment and in the UK are expected to have received clinical pathology (CPA) accreditation, or to be working towards accreditation.8

A couple considering pre-implantation diagnosis must receive genetic counselling to confirm that they have a good understanding of the nature of their particular condition, how it may affect their offspring and that they are aware of alternative reproductive options. If one of the couple carries a lethal or debilitating progressive disorder, they need to consider the welfare, and make arrangements for the care, of any child born. The need to take account of the welfare of the child and any existing children is a statutory requirement of the HFEA.8 Couples need to fully understand the processes involved in pre-implantation diagnosis in terms of stimulation and collection of oocytes, fertilisation, development of embryos and the biopsy process, and the risks of superovulation including OHSS and multiple pregnancy. Couples should be aware of the large attrition in numbers of embryos that may be available from beginning a cycle to those that may be suitable for replacement after pre-implantation diagnosis, the predominant factor in the likely success in terms of pregnancy. They also should be aware of the risk of misdiagnosis after pre-implantation diagnosis and should consider whether they would accept confirmatory prenatal diagnosis.

Pre-implantation diagnosis is a major undertaking for any couple, and the psychological, medical and financial costs are considerable. A single cycle might cost anything from £4000–£7000 with a likely ‘take-home baby’ rate not much above 20% wherever this is undertaken.9 In our practice, approximately half of UK patients are likely obtain NHS funding for at least one cycle. Funding of pre-implantation diagnosis is not one of the issues considered in the recently published NICE fertility guideline as this service appropriately falls within the consideration and funding of genetic services10 ( In some areas, specialist local consortia are being established to advise primary care trusts about the appropriateness of cases for NHS funding.

Of 275 cycles performed for 149 couples in our unit, the cumulative clinical pregnancy rate per couple treated was 36%; (livebirth rate 20% per cycle started). On average, the time interval from referral to consultation is six months. Audit over a 19-month period revealed that of 158 couples referred, only one third started treatment. Most dropped out having established what PGD entailed. Thus, less than 10% of those referred originally went on with treatment and achieved a clinical pregnancy.11

Clinical procedures and embryology

Although patients seeking pre-implantation diagnosis are generally fertile, assisted reproduction techniques are required. Ovarian stimulation, monitoring and egg collection are identical to those used for assisted reproductive techniques for alleviating subfertility, but because the number of eggs (and hence embryos) available is critical to success, regimes tend to be more aggressive. Oocytes may be fertilised using in vitro fertilisation where the sperm quality is deemed adequate. Intracytoplasmic sperm injection must be used when the diagnostic test requires the use of PCR, as the presence of additional sperm buried in the zona pellucida may lead to contamination of the PCR reactions with paternal DNA thus risking misdiagnosis.

For whom may pre-implantation diagnosis be suitable?

Once a biopsy is performed, the type of test applied to the cell depends on the genetic condition being tested. It is offered for three main categories of disease (Table 1):

Table 1.  Common indications for pre-implantation genetic diagnosis arranged by mode of inheritance.
Single gene disorders
Autosomal dominantHuntington disease
Myotonic dystrophy
Charcot-Marie-Tooth 1a and 2a
Osteogenesis imperfecta I and IV
Stickler syndrome
Tuberous sclerosis
Marfan syndrome
Autosomal recessiveCystic fibrosis
Spinal muscular atrophy
Sickle cell disease
Epidermolysis bullosa
Gaucher disease
Tay–Sachs disease
Sex-linked dominantFragile X syndrome
Oro-facial-digital syndrome type 1
Sex-linked recessiveDuchenne muscular dystrophy
Spinal and bulbar atrophy atrophy
Hunter syndrome
Chromosomal disorders
Structural aberrationsReciprocal translocation
Robertsonian translocation
Numerical aberrations47 XXY (Klinefelter Syndrome)
47 XYY
Sex chromosomal mosaicism
Male meiotic abnormalities
  • (i) to determine the sex of an embryo for a X-linked condition whose specific gene defect at a molecular level is unknown, highly variable or unsuitable for testing on single cells;

  • (ii) to identify single gene disorders, which may be recessive or dominant, where the molecular abnormality is tested following amplification by PCR of DNA extracted from single cells;

  • (iii) to detect a variety of chromosomal arrangements such as translocations, inversions, deletions or insertions on interphase nuclei using fluorescence in situ hybridisation.

X-linked disorders

Where a specific test at the single-cell level is not available, couples may opt to have sex selection of the embryos using fluorescence in situ hybridisation in order to replace female embryos only thereby preventing the condition.12 This does not eliminate the disease from the family in the future as half of these female embryos will be carriers. It is often assumed that since on average half the embryos will be of either sex, the use of sex selection gives a 50:50 chance of being able to select an unaffected embryo from the cohort. In practice, the odds are worse than this, as some embryos may be unsuitable for replacement because the test has revealed aneuploidy or an uncertain diagnosis due to fluorescence in situ hybridisation failure.13 Also, the fact that half the non-transferred male embryos could be normal may raise ethical objections. As mentioned, skewing the spermatozoal population by using fluorescence activated cell sorting to increase the number of X-bearing spermatozoa before in vitro fertilisation may be used to increase the number of available unaffected embryos.7 In general, this test is robust, as misdiagnosis would have to involve two errors; loss of a Y signal and gain of an X. Despite this, one fluorescence in situ hybridisation misdiagnosis for sexing has been reported (Fig. 2).9

Figure 2.

Fluorescence in situ hybridization using target-specific DNA probes labelled with different fluorochromes is used to identify the sex of embryos for X-linked diseases. This picture depicts two nuclei (left) from a single blastomere taken from the corresponding embryo (right). The two nuclei that have been hybridised with probes that are complimentary to sequences on the sex chromosomes X (green), Y (red) and chromosome 18 (blue) which is used as a control for ploidy. A nucleus from the blastomere of a normal female embryo (top right) has two green and two blue signals, whereas a nucleus from a normal male (bottom right) has one red, one green and two blue signals.

Chromosomal rearrangements

Chromosomal rearrangements are a well-recognised cause of reproductive failure.14 Reciprocal translocations (where there is an exchange of terminal segments from two different chromosomes) occur in about 1 in every 500 of the general population. They are more common in infertile couples (0.6%), men with oligoasthenoteratospermia (3%) and in couples with repeated in vitro fertilisation failures (3.2%).15,16 In fertile couples with three or more consecutive first trimester miscarriages, they are present in almost 10% of cases. A carrier of a balanced reciprocal translocation is phenotypically normal but due to abnormal segregation at meiosis, 50% to 70% of their gametes are likely to be unbalanced. Because each translocation is unique, and there are 30 possible unbalanced segregation products and only 2 which are phenotypically normal (balanced and normal), the specific effects of the translocation such as miscarriage, affected liveborn risk and reproductive failure will be particular to the specific chromosomes involved and their breakpoints.17

In order to perform pre-implantation diagnosis for reciprocal translocations, fluorescent probes specific for the subtelomeric regions of the translocated segments and the centromere regions of centric segments are hybridised to the interphase nucleus of the biopsied cell (Fig. 3). For Robertsonian translocations (where two acrocentric chromosomes fuse to give one derivative chromosome), chromosome enumerator probes, which can be chosen to bind at any point on the long arm of the chromosome that is involved in the translocation, are used to count the chromosomes in the nuclei of the cells.18

Figure 3.

A cartoon showing 3 of the 32 possible chromosome rearrangements and corresponding fluorescence in situ hybridization patterns for a reciprocal translocation of 46XXt(1;6) showing an exchange of material between the short arms of chromosome 1 (yellow) and 6 (blue). Fluorescent DNA probes specific for the chromosome telomere regions (red and green) and one centromeric region (blue) have been hybridised to the chromosomes. The pattern for normal or balanced chromosomes is indistinguishable. The top pattern is for the balanced form and the lower patterns are two possible unbalanced forms.

The most common indication for pre-implantation diagnosis for translocations is to prevent repeated miscarriage. Other causes of miscarriage such as antiphospholipid syndrome or intrauterine abnormalities first must be excluded as they cannot be prevented using pre-implantation diagnosis but may be present in addition to the chromosome rearrangements. The chance of a live birth following spontaneous conception in the next pregnancy even after three first trimester losses is in excess of 65% and still 62% after four losses.19 So the decision to use pre-implantation diagnosis, where the liveborn chance is only of the order of 20%,9 is debatable and needs to take account of other factors (e.g. psychological consequences of loss and subfertility).

Single gene disorders

As the amount of DNA available for diagnosis in a single cell biopsy is tiny, a direct test cannot be performed without amplification of the DNA using PCR to a level where the specific mutation can be detected.1 At this level, PCR may not be wholly reliable such that occasionally, sequences may fail to amplify. This could result in a heterozygous embryo being typed as homozygous affected and thus excluded from the cohort of embryos available for transfer—false positive.20,21 Equally the classification of a heterozygote as homozygous normal due to absence of the affected allele (false negative) may still be safe in autosomal recessive disorders since there are no adverse phenotypes in carriers, but in autosomal dominant conditions where heterozygotes are affected, problems of undetected allele drop out (ADO) or amplification failure makes the possibility of misdiagnosis too great for a simple test to be acceptable.22 In addition, the sample could become contaminated with extraneous non-embryonic DNA leading to misdiagnosis. The accuracy of the test and the possibility of misdiagnosis (around 4–7%) can be decreased by the use of multiplex fluorescence PCR.23,24 Here simultaneous amplification of two or more loci, one containing the mutation and one or more containing informative polymorphic markers in close proximity to the mutation (acting as a mini-fingerprint), confirms the embryonic origin of the DNA. The enhanced sensitivity of fluorescence PCR makes ADO less frequent, and because contamination by rogue DNA can be detected as it can be distinguished from DNA of embryonic origin, the risk of misdiagnosis is reduced.

Regulation of pre-implantation genetic diagnosis

There are substantial international differences in the regulation of pre-implantation diagnosis. In the United Kingdom, all assisted reproduction techniques including pre-implantation diagnosis are regulated by the HFEA under the terms of the Human Fertilisation and Embryology Act (1990). According to the Act and the 6th Code of Practice, an embryo may only be used for pre-implantation diagnosis or for research to develop diagnostic methods under licence from the HFEA.8 The HFEA provides reassurance to the public that pre-implantation diagnosis is being undertaken only for serious genetic conditions and, not simply for social reasons. Pre-implantation diagnosis has only recently been allowed in France in three specified centres. It is disallowed in several countries including Austria, Germany and Switzerland. Recent legislation proposed in Italy has caused consternation, as its restriction will outlaw pre-implantation testing. There is no federal regulation of pre-implantation diagnosis in the United States.

There are understandable concerns regarding the safety of pre-implantation diagnosis. The long term effects of removal of one or more cells at cleavage stage, which will decrease the cellular mass of the embryo, are unknown. Although animal experiments suggest that this does not adversely affect in vitro and in vivo development, there are few good follow up studies of embryo biopsy in humans. In one small study of 123 deliveries, the complications of pregnancy, birthweight and length of pregnancy were found to be similar to intracytoplasmic sperm injection. Of 109 infants born following polar body biopsy, six birth defects—two major (an amniotic band and neonatal seizures), and four minor were reported.25 An increased multiple monozygotic pregnancy rate has also been reported (27–31%) and triplets have occurred following two-embryo transfer, although whether this is increased significantly is still under investigation. Success of pre-implantation diagnosis seems to be universally similar with quoted rates clinical pregnancy of 19% per oocyte retrieval and 23% per embryo transfer. So far the reported misdiagnosis rate is higher in the PCR group (3.4%) than in the fluorescence in situ hybridisation group (0.9%).9

In Europe, the ESHRE PGD consortium was formed in 1997 to undertake a long term study of the efficacy, safety and clinical outcome of pre-implantation diagnosis. It reports its findings annually at the ESHRE meeting ( In 2002, an international society, the Preimplantation Genetic Diagnosis International Society (PGDIS) was formed to encourage and co-ordinate research, education and training in pre-implantation diagnosis on a more international basis.

New advances in pre-implantation genetic testing

Pre-implantation genetic screening

The use of pre-implantation diagnosis to prevent transmission of serious genetic disease has largely been overtaken in volume by use of embryo biopsy to improve in vitro fertilisation outcome for subfertile couples.26 The application of fluorescence in situ hybridisation as well as other more sophisticated techniques has been extended to detect embryos that contain major sporadic chromosomal or age-related aneuploidies that may result in failure of implantation or spontaneous miscarriage, and to remove them from the cohort available for transfer. This technique is variously called pre-implantation genetic screening by the HFEA in the UK, or aneuploidy screening (PGD-AS) in other parts of Europe and by the ESHRE consortium. In the USA, it tends to be included under the overall banner of pre-implantation diagnosis, which inflates the number of pre-implantation diagnosis cycles that have been reported from there.

During pre-implantation genetic screening, individual embryos are biopsied, and single blastomeres are examined using between 5 and 14 fluorescence in situ hybridisation probes (most commonly X, Y, 13, 15, 16, 17, 18, 21 and 22).27 The technique has so far been advocated for three main groups.

(a) Recurrent miscarriage. Recurrent miscarriage affects about 1% of couples of reproductive age.28 Although chromosomal abnormalities are implicated in up to 60% of first trimester miscarriages, less than 5% of couples with two or more miscarriages carry a balanced chromosomal rearrangement.29De novo numerical abnormalities (in particular, monosomy X and autosomal trisomies involving chromosomes 16, 21 and 22 and to a lesser extent chromosomes 13 and 18) account for the vast majority of spontaneous miscarriages caused by chromosomal aberrations. Thus, screening embryos with fluorescence in situ hybridisation analysis using probes specific for these chromosomes could detect approximately 70% of the chromosomal abnormalities found in spontaneous miscarriages and provide a means of improving reproductive outcome in affected couples.

(b) Poor prognosis in vitro fertilisation patients. Advanced maternal age and recurrent implantation failures (defined as three or more failed in vitro fertilisation attempts) are poor prognostic factors for successful in vitro fertilisation treatment.30 Studies of pre-implantation embryos in women of advanced maternal age suggest that more than half have aneuploidies detected by fluorescence in situ hybridisation including trisomies involving chromosomes 16 and 21. Likewise, up to 75% of embryos from patients with repeated in vitro fertilisation failures may be chromosomally abnormal, the most common abnormalities encountered being aneuploidies of chromosomes 13, 16, 21 and 22.31 Because good morphology and developmental stage cannot reliably predict chromosomal normality of early human embryos, the use of fluorescence in situ hybridisation to exclude chromosomally abnormal embryos has been advocated to improve the success rate of in vitro fertilisation in those poor prognosis patients particularly in light of the current tendency to limit the number of transferred embryos to avoid multiple pregnancies.

(c) Severe male factor infertility. Although intracytoplasmic sperm injection has dramatically changed the outlook for conception in couples with severe male factor infertility, it appears that among those patients there are various underlying genetic abnormalities that may be involved in the aetiology of their infertility. For example, patients with non-obstructive azoospermia or severe oligospermia may have microdeletions in the long arm of the Y-chromosomes.32 In addition, intracytoplasmic sperm injection cycles aimed at the treatment of severe male factor is associated with an increased risk of de novo chromosomal aberrations especially aneuplodies involving the sex chromosomes.33 These may be secondary to increased aneuploidy rates in the sperm used.34 Hence, applying the appropriate fluorescence in situ hybridisation probes to those intracytoplasmic sperm injection-generated embryos may reduce transfer of such embryos.

Limitations and drawbacks of pre-implantation genetic screening

Despite a decade of experience, over 3000 treatment cycles of pre-implantation genetic screening reported and over 500 babies born, data from randomised controlled trials demonstrating the efficacy of the technique over less invasive alternative therapies are lacking. Furthermore, there are no studies clearly showing improvement in live birth rates per stimulated cycle with the use of pre-implantation genetic screening, particularly when taking into account the advantage that additional frozen embryos may confer. To date, most studies have reported their results in terms of implantation and clinical pregnancy rates per transfer, falsely inflating the reported success rate by excluding cycles where embryo transfer was not performed because no normal embryos were available for transfer (usually between 20% and 65% of cycles). Only one recent multicentre study of pre-implantation genetic screening showed a significant twofold increase of implantation from 10.2% to 22.5% based on clinical pregnancy, and a decrease in the miscarriage rate from 27.5% per patient in the control group to 14.3% in the group undergoing PGD-AS testing. Although clinically impressive, this difference was not statistically significant.35,36

False-positive and false-negative fluorescence in situ hybridisation results (due to hybridisation failure, weak or overlapping signals and cross-hybridisation), estimated in various studies to be between 5% and 12%, can lead to chromosomally normal embryos being excluded from the cohort suitable for transfer. In addition, the fact that mosaicism is a natural event in early human embryos, even in morphologically normal good quality embryos, undermines the accuracy of pre-implantation genetic screening as a screening test for embryo euploidy since the cell removed and tested may not be representative of the entire embryo.37 The overall outcome of these pitfalls is that 12–20% of normal embryos may be excluded from transfer, placing women with a low embryo yield at a significant risk of not having embryos transferred at all. Finally, even if a good number of biopsied embryos have been found ‘normal’, women failing to conceive after the fresh transfer are disadvantaged as the surplus ‘normal’ embryos do not have the same potential for survival after cryopreservation and thawing compared with intact embryos without a zona breach.

There is a clear need for randomised controlled trials examining the outcome of pre-implantation genetic screening in terms of live birth per cycle started in order to define the groups for whom this technique may be beneficial. This is an expensive and so far unproven test that is taken up by women desperate to gain a pregnancy when, for some it might actually decrease their chances of pregnancy by reducing the cohort of normal embryos available for transfer.

Social sex selection

The use of pre-implantation diagnosis for sex selection to prevent X-linked disorders has been extended to the selection of a child of a particular gender. This may be either by preference where selection is predominantly for one sex when only one child is allowed, or where male offspring are favoured over female for cultural or economic reasons. Alternatively, it may be performed for ‘family balancing’ where there are already one or more children of the same sex in the family and there is a desire for a child of the other sex. In the UK, after a recent second public consultation, the HFEA has reaffirmed its original position in opposing any form of sex selection for non-medical reasons.38

This is not the case in many other countries in the world where sex selection is viewed as acceptable and sex selection by fluorescence in situ hybridisation is permitted as an alternative to prenatal diagnosis with a predictable bias towards male children.39 However, as it is still an expensive technique that requires substantial expertise and equipment, it is not widely available and thus has little chance of altering the sex ratio compared with abortion or feticide, which is still practised in a number of countries even where the practice is outlawed.

Pre-implantation HLA typing

Pre-implantation HLA matching has been used to ensure that an embryo to be replaced after in vitro fertilisation and pre-implantation diagnosis will be a suitable tissue match for an affected sibling, where it is intended that cord blood will be harvested at birth for use in bone marrow cell therapy.40 The first such case was Adam Nash, born following pre-implantation diagnosis not only to exclude his being affected by Fanconi anaemia, but that he should be a suitable match for his sister Molly who suffered from the condition and required a bone marrow transplant.41 In these situations, a large number of embryos are likely to be discarded as only 3 in 16 embryos will both be unaffected by the recessive disorder and be a full HLA match. Coupled with this is a 20–25% success rate of pregnancy which may be worse depending on the woman's age. Several pre-implantation diagnosis cycles might have to be performed to achieve a match let alone a successful pregnancy. Pre-implantation HLA genotyping in combination with pre-implantation diagnosis for causative genes has also been performed for thalassaemia, hyperimmunoglobulin M syndrome, X-linked adrenoleucodystrophy and Wiskott Aldrich syndrome.2 HLA typing without pre-implantation diagnosis in the absence of a high risk genetic transmissible disease is even more controversial. Here there may be no pre-existing genetic condition to be avoided, but the pre-implantation diagnosis performed with the sole objective of pre-selecting HLA matched progeny for cell therapy treatment of siblings. Although the likelihood of match is higher (one in four embryos is likely to be a suitable match), the fact that 75% of the created essentially normal embryos will be unsuitable and discarded leads to substantial ethical disquiet as the embryos are being typed and conceived for no other reason than to be a match for an affected brother or sister.42 Issues of consent and protection of the children's autonomy are raised, especially should the cord blood fail to yield sufficient stem cells or if the cell therapy should fail.

Future developments

Rapid and efficient analysis of all 46 chromosomes is the ideal as some embryos diagnosed as normal can still be abnormal for other chromosomes not currently analysed. So far the most promising method is comparative genomic hybridisation where molecular techniques are used to perform a quantitative analysis of the whole genome, although polyploidy and balanced translocations cannot be detected.43 At present, this technique requires two or three days for diagnosis, and is therefore unsuitable for routine use on cleavage stage embryos or blastocysts which need to be cryopreserved until the diagnosis has been made. Several ongoing pregnancies and the birth of one healthy child have been reported, but at present the delay in diagnosis, and the damage caused by the freezing and thawing of biopsied embryos probably outweigh the benefits of comparative genomic hybridisation. Faster more robust techniques, such as DNA microarrays, may obviate the need for embryo freezing. This is a method of molecular analysis primarily used for gene expression analysis. However, it could also be applied to routine pre-implantation diagnosis to screen for mutations in any one gene, or screening several genes for several mutations.44 Microarrays might also be useful in pre-implantation diagnosis of specific conditions that are genetically heterogeneous and for which there are few common mutations such as Duchenne muscular dystrophy, where it could provide a generic testing procedure applicable to all patients carrying the gene. The ability to screen mutations in one gene or several mutations for several genes would allow embryos to be tested for serious susceptibility traits loci, such as the breast cancer gene (BRCA1). Finally, appropriately constructed microarrays might replace the need for metaphase spreads now used to assess chromosome imbalance during comparative genomic hybridisation.


Pre-implantation diagnosis combines the technology of assisted reproduction and molecular and cytogenetics in order to perform a very early form of prenatal diagnosis. Preliminary information and results so far suggest that it is probably safe and reliable. Although there is little evidence of impairment of embryonic development, there is no good evidence for estimating its long term safety and there is still a small but significant risk of misdiagnosis. Thus, quality standards should be as rigorous as for other forms of prenatal testing.

Rapid advances in molecular genetics will allow efficient examination of the entire chromosomes of the embryo. More couples are likely to seek using the technology in order to improve their chances of successful in vitro fertilisation. It is important that the use of pre-implantation diagnosis is strictly regulated for medical purposes and prohibited for eugenic selection.

Useful reviews

Braude P, Pickering S, Flinter F, Ogilvie CM. Preimplantation genetic diagnosis. Nat Rev Genet 2002;3:941–953.

Munne S. Preimplantation genetic diagnosis and human implantation—a review. Placenta 2003;24(Suppl B):S70–S76.

Sermon K, Van Steirteghem A, Liebaers I. Preimplantation genetic diagnosis. Lancet 2004;363:1633–1641.