Preimplantation genetic diagnosis: general principles and focus on haemoglobinopathies

Summary
According to acceptable practices in each society, at-risk couples can be offered reproductive choice either for the avoidance of an affected child, or to support preparation for the possible birth of a severely affected child. For couples who prefer to have only healthy children, reproductive choices can include, changing partner choice; remaining childless, opting for gamete donation or adoption, but more commonly couples at risk choose to avoid the birth of an affected child in early pregnancy by undergoing conventional prenatal diagnosis or before pregnancy by choosing pre-implantation genetic diagnosis (PGD). Although a well-established and highly reliable procedure, conventional prenatal diagnosis has a few disadvantages. It requires invasive acquisition of fetal material (through amniocentesis or trophoblast sampling) which may be associated with a small risk to the pregnancy. Moreover, if the pregnancy is diagnosed as affected, the couple faces the difficult choice of pregnancy termination, with associated ethical and psychological implications. The mian advantage of PGD compared to conventional prenatal diagnosis is that pregnancy termination is precluded. Thus for couples at high-risk of transmitting an inherited disorder, (either a monogenic disease or a structural chromosomal abnormality), PGD supports the initiation of a pregnancy unaffected for the familial disorder. PGD is applied to diagnose the specified genetic condition in cells biopsied from oocytes/zygotes or embryos obtained in vitro through assisted reproductive techniques (ART). Following analysis, only those embryos unaffected for the disease under consideration are transfered to the uterus. The major advantage over conventional prenatal diagnosis is that PGD avoids the need to terminate affected pregnancies. However, it is not an easy solution. Despite many advances, PGD remains for the most part a technically challenging, multi-step and labour-intensive procedure, requiring close collaboration between many specialists. Although the genetic analysis can be stringently optimized, limitations include the requirement to involve ART, even if the couple are fertile, the high cost of a PGD cycle and the fact that pregnancy and birth rates rarely surpass 30-35%. Thus, PGD is generally reserved for couples with an unsuccessful reproductive history, including fertility problems (in which case they require ART anyway) or an objection to pregnancy termination (either ethical/moral, or following experience of a previous termination for medical reasons).

A short definition
Preimplantation genetic diagnosis (PGD) may be considered an early form of prenatal diagnosis. The major advantage compared to conventional prenatal diagnosis is the initiation of unaffected pregnancies, which precludes pregnancy termination if the couple prefers to have only healthy children. It is suitable for couples at high-risk of transmitting an inherited disorder to their offspring, either a specific monogenic disease(s) or structural chromosomal abnormality. Its goal is to diagnose the specified genetic condition in cells biopsied from oocytes/zygotes or embryos obtained through assisted reproductive techniques (ART), and following analysis, to transfer to the uterus only those embryos identified as unaffected for the disease under consideration to establish a healthy pregnancy.

Introduction and impact
PGD gives couples at risk of passing a genetic disorder to their children, the chance of establishing a healthy pregnancy and having a healthy child. The first reports for clinical PGD were over 20 years ago, in 1990. PGD combines ART and genetic diagnosis. Before the couple can be accepted for a PGD treatment they must have thorough evaluation for their reproductive and genetic status, along with comprehensive counseling on all aspects of ART and on all possible genetic diagnosis outcomes,. Once informed on all aspects of the treatment they should provide consent. The couple must undergo ART even if they have no fertility issues. ART in PGD is identical to standard infertility treatments, except for the step of embryo biopsy (and genetic analysis). The oocytes and embryos created in vitro are biopsied to provide single cells representing each oocyte/embryo (the latter depending on the stage of biopsy – see section ART and zygote or embryo biopsy). These are then subjected to stringent genetic testing to determine the genetic status of the embryo(s) relative to the genetic disorder or chromosomal abnormality that runs in the family. Often customized genetic tests have to be developed which are specific for each family. Good quality embryos with an unaffected genetic status for the familial condition can be selectively transferred to the womb of the candidate mother. Since PGD is a multi-stage and relatively complex procedure overall, it is not necessarily a replacement for conventional prenatal diagnosis. However, it is most appropriate for couples with a risk of passing on a genetic disorder to their progeny, especially if in addition: a) They require ART due to fertility problems; b) They have a history of pregnancy termination following conventional prenatal diagnosis; c) They have ethical or religious dilemmas with regards to pregnancy termination, . The procedure of PGD can be applied today for any single gene disorder or chromosomal rearrangement as long as: a) The disorder has been clearly characterized and evaluated by a clinical geneticist; b) A genetic diagnosis report is available, in other words, the fault at the DNA/chromosome level that is associated with the genetic problem in the family has been identified. Generally PGD is usually applied to preclude severe monogenic or chromosomal disorders and for these applications there is usually little ethical debate. However, some of the newer uses that have emerged, such as PGD for autosomal dominant late-onset disorders, cancer predisposition syndromes, PGD for histocompatibility (HLA) typing and mitochondrial DNA (mtDNA) mutations, are associated with greater ethical controversy and ethical questions. Despite the complexity of the PGD proceduer, according to 11 years of data collections from the European Society of Human Reproduction and Embryology (ESHRE), between 1997-2008, PGD has been successfully applied in >6000 cycles to exclude monogenic disorders and >4000 for familial structural abnormalities, with approximately 2000 healthy babies born to date, ,. Not unexpectedly cycles to exclude cases affected with haemoglobin disorders are amongst the most frequently applied. Overall PGD represents a valuable reproductive alternative for couples at risk for transmitting a genetic condition who wish to avoid the need to terminate affected pregnancies.

Technical aspects of PGD
PGD represents a means to avoid the termination of affected pregnancies, through the identification and selective transfer of unaffected embryos created by assisted reproductive techniques (ART). PGD is now a widely established reproductive alternative for couples with a high-risk of transmitting an inherited disorder, although regulatory frameworks may vary between countries. A technically challenging, multi-step procedure, PGD requires the close collaboration of experts from several medical specialities, including gynaecologists experienced in ART, embryologists skilled in embryo biopsy and geneticists skilled in molecular technologies. PGD is generally reserved for couples with a difficult or unsuccessful reproductive history, including fertility problems (in which case they will undergo ART procedures anyway) or an objection to pregnancy termination (either ethical/moral objection, or for medical reasons after experience of a previous termination). Since the first clinical cycles applied over 20 years ago, PGD is currently offered in a substantial number of specialized centres throughout the world (http://www.eshre.com) for a huge range of monogenic diseases (as long as there is a definitive molecular diagnosis and/or clear linkage defined within a family), as well as structural chromosome abnormalities.



The many stages in PGD include evaluation and counselling of the couple on aspects of genetics and reproductive potential, application of all stages of ART; biopsy of genetic material representing each embryo, the genetic analysis, and, if implantation occurs subsequent to embryo transfer, follow-up of the pregnancy and baby (or babies) delivered (Figure 1).

ART and zygote or embryo biopsy
ART procedures for PGD are the same as for infertility treatment, (with the exception of the step for oocyte/zygote or embryo biopsy), irrespective of whether the couple is fertile or not. Biopsy of material for genetic analysis may be performed at various stages post-fertilization, including polar body biopsy on the first day post-insemination, blastomere biopsy (1-2 cells) from cleavage-stage embryos on the third day post-insemination, or 5-10 trophectoderm cells from blastocysts on the fifth day post-insemination. Each stage has relative advantages and limitations (reviewed in reference 4), but most PGD cycles reported to date have been based on blastomere biopsy. It is the role of the gynaecologists in collaboration with the embryologists to select the most appropriate option for each PGD cycle/case.

Genetic analysis for monogenic diseases
Common to all types of biopsy is the very limited quantity of sample available for genetic analysis, which is usually only a single cell. The limited sample is considered the most technically challenging aspect of PGD, often compounded by the sub-optimal quality of the embryo and/or embryo cell biopsied. For monogenic diseases PGD protocols always require an initial amplification step, usually based on the polymerase chain reaction (PCR), although some protocols use whole genome amplification (WGA). (see below). Either way, prerequisites of PGD protocols include: speed (to produce a result within 24-72 hours), sensitivity (to successfully derive the genotype of a single cell), robustness (to always produce a result when the sample quality is good) and accuracy, which must approach the ideal 100%, so that affected embryos are never transferred. The optimal conditions and strategies for performing PGD are described in the ESHRE Best Practice Guidelines.

All PGD protocols require a stringent work-up prior to clinical application, to address innate limitations of single-cell PCR. These limitations include total PCR failure, failure to detect both alleles (allelic drop-out, or ADO), and sample contamination. Contamination can potentially occur at either the stage of embryology biopsy or genetic analysis during the PGD, but can be minimized by using stringent laboratory procedures, such as physically separate areas for pre-PCR, PCR and finally post-PCR steps. PCR set-up should be done in dedicated UV-treated hoods, using exclusive UV-treated equipment, entirely separated from any post-PCR processing, and with operators wearing appropriate attire and employing correct handling procedures. Overall, PCR failure, although undesirable, will not lead to an unacceptable misdiagnosis. On the other hand, ADO and contamination may lead to serious misdiagnosis. Besides the high diagnostic value of a PGD genetic analysis, it must address the limited time-frame (24-48 hours), to permit timely embryo transfer.

Once a couple has consented to enter the PGD programme, an appropriate protocol for genetic analysis has to be developed. The first step involves in silico protocol design, through the use of electronic databases to locate the DNA sequences of the causative gene and the area around it. Protocols can directly analyse the mutation as well as closely linked genetic markers (usually short tandem repeats, also known as STRs) that are identified at the in silico stage. Primers are designed to facilitate optimal amplification in a single multiplex reaction so that both the mutation region(s) and the STRs can be co-analysed during the PGD, the latter to support linkage analysis. This increases the number of loci which can define the genetic status of each cell. Once the primers have been designed and ordered, the next step is to test the multiplex PCR protocol on dilutions of DNA, confirming that a result can be obtained even from low amounts of DNA (similar to the amount in a single cell). The final step is to confirm that the protocol works on single cells, usually using lymphocytes isolated from the blood of the prospective parent that carries the mutation, as a model. The aim is to achieve success of >90% PCR amplification success and <10% ADO for each loci (usually at least 4-5) in each PGD test. Ideally over 50 single cells are tested to validate the protocol, according to the ESHRE PGD consortium guidelines.

An optimized multiplex PGD-PCR protocol facilitates analysis of preferably no less than 4 linked markers (two upstream and two downstream to the causative gene) across the disease-associated locus, addressing the pitfalls of ADO and monitoring contamination. If ADO occurs at one genetic locus then an accurate result may be derived through analysis of the remaining loci. For contamination, the detection of spurious STR allele sizes not carried by the parents indicate contamination and thus reduce the validity of the genetic result. In addition stringent laboratory procedures must be applied at all stages during PGD to ensure optimal efficiency and accuracy of the PGD result, supporting the transfer of unaffected embryos. For some of the more common genetic diseases for which PGD is more frequently applied, including the haemoglobinopathies, several generic protocols have been developed, which preclude the design and validation of patient-specific protocols for every case. (see section PGD for the haemoglobinopathies).

The disadvantages of testing single cells with PCR-based protocols can potentially be overcome by using an initial step of whole-genome amplification (WGA) on the single (minimal) cell samples prior to a genotyping step. The genotyping step can then be applied under standard laboratory conditions since the WGA should produce ample amounts of DNA and the sample can be treated as would a typical genomic DNA sample. Some WGA protocols are PCR-based e.g. Degenerated Oligonoucleotide Primed PCR (DOP-PCR) or Primer Extension PCR (ΡΕΡ), although early reports of these methods described limited success. Subsequent PCR-based methods involve adaptor-mediated amplification of a library of short, overlapping amplimers. Other developments include the use of an alternative polymerase (bacteriophage φ29) which is based on the principle of multiple displacement amplification (MDA). MDA potentially provides more efficient and faster DNA amplification from single cells. Overall most WGA protocols still have major limitations when applied to single cells. They usually do not amplify the entire genome of a single cell with homogeneous efficiency, and generally high levels of ADO post-WGA may be observed (up to 30%) at various (usually random) genomic regions. Consequently, although WGA potentially supports more generic and highly multiplexed downstream protocols for the PGD, such as Preimplantation Genetic Haplotyping (PGH), it may only be worthwhile when PGD protocols need to address analysis of diseases associated with large genes and many potential mutations (for example Cystic Fibrosis or Duchenne Muscular Dystrophy); with respect to PGD applications for haemoglobinopathies, WGA may be useful when selecting donor siblings to facilitate bone-marrow transplant (see below).

PGD for the haemoglobinopathies
The haemoglobin disorders are the most common autosomal recessive disorder worldwide and thus it is logical that it is one of the diseases for which PGD is most commonly applied (according to the ESHRE data collections). As with classic PND, PGD applications tend to be restricted to the haemoglobinopathies with severe clinical expression (see http://www.thalassaemia.org.cy/wordpress/wp-content/uploads/2012/12/PREVENTION_BOOK_FINAL.pdf), mainly β-thalassaemia major, possibly the sickle cell syndromes and Hb Bart’s or HbH hydrops foetalis. According to the annual PGD data collections published by the European Society of Human Reproduction and Embryology (ESHRE), PGD for haemoglobinopathies is one of the most common applications for monogenic diseases, with over 700 cycles reported to date. The first PGD cycles for β-thalassaemia were reported in 1998. Diagnostic strategies were based on analysis of polar bodies and genotyping was achieved for the presence/absence of the globin gene mutation(s), using nested-PCR with restriction enzyme analysis. Since then, methods and technologies have evolved. Although almost 200 clinically relevant mutations have been described associated with the β-haemoglobinopathies, the small size of the gene and clustering of most mutations within the gene means many of the more recent PGD protocols described are generic and applicable for a wide range of genotype interactions. This precludes the need to develop patient-specific protocols each time, and is particularly appropriate for the haemoglobinopathies since PGD centres can likely expect a relatively high number of requests for cycles. More recent methods described for PGD of haemoglobinopathies involve minisequencing or real-time PCR, along with multiplexed analysis of polymorphic sites for linkage analysis,.

PGD to preclude the severe forms of α-thalassaemia are more relevant for populations of Southeast Asia and China, where deletions of both functional HBA genes from the α-globin gene cluster are common (e.g. Hb Bart’s hydrops foetalis). Several protocols for PGD applied to cases originating from Southeast Asia and China have been described, based on either multiplex fluorescent gap PCR or linkage analysis of polymorphic sites within the disease-associated locus. Recently a generic protocol based on multiplex linkage analysis of polymorphic STRs located within the α-globin gene cluster was described for PGD to preclude severe forms of HbH disease and HbH-Bart’s disease encountered in the Mediterranean populations, ,.

PGD for donor sibling selection and haematopoietic stem cell transplantation
For carrier parents who already have an affected child, PGD can be applied for human leukocyte antigen (HLA) testing in order to obtain another child who can be a source of HLA-matched haematopoietic stem cells to support transplantation to treat the affected sibling(s). PGD-HLA was first described in 2001 to treat a child with Fanconi anemia. Generally for HLA-typing in families who also have a hereditary haematological condition, besides the HLA typing, the PGD analysis also has to ensure the selection of unaffected embryo(s). Thus PGD-HLA protocols involve the co-analysis of many amplicons both across the 3.6Mb of the HLA locus on chromosome 6p21, and the HBB locus on chromosome 11p15.5. Generally PCR-based assays are capable of co-amplifying up to around 20 amplicons in a single multiplex reaction (if well designed and optimized). WGA could also be an appropriate initial step in protocols for PGD-HLA.

PGD-HLA is potentially ethically controversial for donor-sibling selection and has been associated with phrases such as “designer babies”. However, the selection of a histocompatible sibling to facilitate a bone marrow transplant in a thalassaemia major patient is clearly for the benefit of a patient and the sibling-donors are clearly valued by all members of the family with an affected child. In fact PGD-HLA to save a child is considered acceptable in many countries and many of the ethical controversies have been resolved. On the negative side is that the chance of ultimate success i.e. the birth of an unaffected histocompatible baby, is very limited in practice by the genetic chances. Firstly the likelihood that two siblings are HLA-matched is only 25%. If this is combined with the chance that only 75% of matched embryos are unaffected for the haemoglobinopathy, this means that potentially only 18.8% of all embryos fertilized will be genetically suitable for transfer in each cycle. Combined with overall success of implantation, pregnancy and delivery rates which are approximately 30%-35%, the overall success rate for PGD-HLA matching rarely surpasses about 10% for any cycle initiated. It is extremely important that all couples should be clearly counselled and informed about this before embarking on this reproductive option. Despite this, it is estimated that over 500 clinical cycles have been performed to date.

Future developments
One major drawback with PGD is that pregnancy rates rarely surpass 30-35%, even for fertile couples. Thus the selection of the best quality embryo(s) for transfer is a key focus of many research efforts, with the aim of improving outcomes of both ART and PGD. This goal is also linked to the trend towards single embryo transfer (to avoid maternal and neonatal risks associated with multiple pregnancies) which means that there is a need to predict the embryo most likely to implant in any PGD (or ART) cycle. Current means to evaluate potential embryo quality involve assessment of embryo morphology and chromosome ploidy, the latter known as preimplantation genetic screening (PGS). However, neither method has proven and robust clinical utility, and with PGS pitfalls include the need for invasive biopsy, and furthermore embryo mosaicism may cause spurious evaluation of the true embryo status,.

The fast emerging technologies of arrays and next-generation sequencing for genome analysis potentially offer generic approaches for simultaneous PGS and high-risk PGD. One such approach is named ‘karyomapping’, and is based on the use of single nucleotide polymorphism (SNP)-arrays for embryo fingerprinting, enabling, through family linkage (or ‘parental support’), analysis of single-gene disorders along with aneuploidy testing. Such SNP-array approaches can also potentially distinguish normal from balanced embryos in translocation cases, identify uniparental disomy and parental origin of abnormalities. However, currently SNP-array analysis of single-cells involves long, complex laboratory protocols, specific instrumentation, and for data analysis, specialized algorithms and software. Efforts are underway to address these limitations, and some progress is evident.

Next-generation sequencing in PGD may facilitate multiple-gene testing, the detection of single nucleotide polymorphisms, copy number variants and chromosomal aneuploidies, as well as epigenetic profiling. Its use at a single-cell level is currently being explored. Besides providing a more holistic approach to PGD, next-generation sequencing in the patients undergoing ART may additionally support more “personalized” procedures.

Methodologies for non-invasive embryo assessment are also under investigation, for example to identify biomarkers of embryo metabolic activity associated with good quality embryos. Production of gametes from stem cells may also become an option, although likely controversial. The aim of all these research efforts is to support improved outcomes, through ensuring optimal sample quality for PGD analyses, and selection of healthy embryos with highest chance to implant.

Closing remarks
During the twenty-plus years over which PGD has been applied in a clinical context, continuous technical improvements have supported progression from what was initially an experimental procedure to a widely acceptable alternative to conventional PND, especially for couples with an unsuccessful or difficult reproductive history. Most of the technical, practical and ethical issues have been addressed with respect to applications related to preventing haemoglobinopathies or PGD-HLA. Today extremely reliable and accurate single-cell genetic diagnostic protocols are available. It is fundamental to understand that PGD is not a simple procedure and that its application requires the highest standards in laboratory and clinical practice, and adherence to strict guidelines (technical and ethical) for all stages of the procedure.