Healthy children without fear

Abstract

Prenatal diagnostics and preimplantation genetic screening are safe and efficient methods to help parents with inherited mutations for severe diseases to have healthy children. Its expansion into adult‐onset diseases and cancer raises challenges for regulation of assisted reproductive technologies and coverage by healthcare systems.

Assisted reproduction has become a standard treatment for prospective parents who cannot conceive children normally because one or both partners are subfertile or infertile owing to congenital birth defects, childhood infection, cancer therapy or accidents. In vitro fertilisation, often in combination with intracytoplasmic sperm injection (ICSI), offers these patients a safe and efficient option to have children. The development of new technologies has further expanded the application of assisted reproductive technologies (ART) to help patients with inherited mutations for severe diseases—such as cystic fibrosis, Huntington or Duchenne muscular dystrophy—who have a high risk that their children would inherit a devastating disease and either die at young age and/or suffer greatly throughout life. Many couples who prefer to become the genetic parents of a disease‐free child can now achieve this by either prenatal diagnosis (PND) and selective termination of pregnancy or preimplantation genetic diagnosis (PGD) and selective transfer of embryos without the disease‐causing mutations.

Whatever their choice, couples or individuals who carry disease‐causing mutations need genetic and reproductive counselling to inform them about the disease and its causes and its recurrence risk and discuss the options for having healthy children. After counselling, couples need to take an autonomous decision to refrain from having children, to adopt, to opt for gamete donation or to use ART with PND or PGD.

Prenatal diagnosis

In the 1960s, it became possible to obtain amniotic fluid for genetic analysis of the embryo's genome. Amniocentesis is commonly performed at the 16^th week of pregnancy when there is sufficient amniotic fluid to take a sample of about 20 ml for cell culture. It was first done blindly, which means that the puncture site was located by external palpation of the uterus. In the 1970s, ultrasound scanning became available, which drastically improved safety as the gynaecologist can see the puncture needle on a monitor. After culturing cells from amniotic fluid, genetic analysis can be carried out. First, this was mainly karyotyping and enzyme assays to diagnose inborn metabolic conditions. Most monogenic conditions could only be studied after DNA‐based diagnostics became available in the 1980s. This started with the use of linked markers, but now direct tests have been developed for thousands of genes and their mutations. The genetic analysis takes about 2 weeks, which means that a decision to terminate pregnancy has to wait to about 18 or 19 weeks into pregnancy, which is very late and causes a lot of anxiety for future parents.

…couples or individuals who carry disease‐causing mutations need genetic and reproductive counselling to inform them about the disease […] and discuss the options for having healthy children.

Chorionic villus sampling (CVS) for PND was introduced in the 1980s. It allows an earlier diagnosis during the first trimester, and pregnancy can be terminated at an earlier stage in an outpatient clinic. First‐trimester termination is much safer for the woman: the maternal death rate is 1 per 100,000 compared with 7–10 per 100,000 in mid‐pregnancy. Given the decided advantage of earlier diagnosis, we would expect more widespread use of CVS. However, it is not as easily performed as amniocentesis, and some centres have had difficulties implementing it. Nonetheless, both procedures are nearly 100% sensitive and relatively safe in experienced hands with procedure‐related losses of about 1 in 400.

Preimplantation genetic diagnosis

Soon after the birth of Louise Brown as the first IVF baby in 1978, scientists and clinicians began to develop methods for selecting embryos in vitro to avoid the need for selective abortion after PND. Couples may also prefer PGD over PND for religious or ethical beliefs on abortion or opt for PGD after one or more unsuccessful rounds of PND. Finally, they may opt for PGD because they require IVF or ICSI anyway owing to infertility.

Since the principle of PGD is based on embryo selection, it requires a relatively high number of embryos to start with. This is achieved by hormonal stimulation of follicle growth, which can cause ovarian hyperstimulation syndrome in about 2% of the women. This condition varies from mild to severe and in very rare cases can lead to death. Moreover, IVF or ICSI and PGD are no guarantee for success. It depends on the woman's age, but on average gives a chance of 25% per treatment cycle that a baby will be born without the disease. After three cycles, the cumulative success rate is about 60%. Furthermore, it is expensive and not always fully reimbursed by national healthcare systems.

Biopsy and molecular analysis for PGD

At day 3, when the embryo consists of about eight cells, one cell is usually taken for genetic analysis. Although this procedure has been practised for about 25 years, it was only recently shown that this is not the optimal time point. Early biopsy significantly impairs the implantation potential of the embryo, whereas biopsy at the blastocyst stage does not 1. An advantage of the latter is that multiple trophectoderm cells are available at day 5 or 6. However, there are concerns that culturing embryos to the blastocyst stage might cause long‐term and transgenerational effects on development and disease risk 2.

It seems that PGD is more acceptable to couples at risk than PND in order to prevent inheriting their mutations.

Preimplantation genetic diagnosis requires methods that work on single cells. First, PCR was used to diagnose autosomal dominant, autosomal recessive and X‐linked monogenic disorders; later, fluorescent in situ hybridisation (FISH) was applied for detecting chromosomal abnormalities. These methods are now being replaced by newer techniques, for example array comparative genomic hybridisation (CGH), which can be carried out on uncultured cellular material. However, array CGH has also disadvantages compared with traditional karyotyping: it cannot distinguish a balanced translocation from a genetically normal, non‐rearranged, chromosomal complement. Low‐level mosaicism is difficult to detect, and triploidy cannot be excluded. Depending on the particular genetic defects, other methods can be used after whole‐genome amplification: microarray comparative hybridisation, genome‐wide SNP analysis, PCR and next‐generation sequencing (NGS). The method of choice for PND is karyomapping, which uses multiple displacement amplification (MDA) followed by genotyping of a large number of polymorphic markers. Genome‐wide karyomapping is highly accurate and can detect any single‐gene defect, or any combination of loci, at the single‐cell level. It has greatly expanded the range of conditions for which PGD can be offered without the need for customised tests 3.

A new, non‐invasive diagnostic method has become available as an alternative for invasive diagnosis. It is based on the analysis of cell‐free foetal DNA (cfDNA) in the maternal plasma, as first described by Dennis Lo in 1997 4. Random genome sequencing not only enables the detection of trisomy 21 and other trisomies, but also inherited chromosomal and monogenic abnormalities. The first tests were introduced for paternally inherited autosomal dominant disorders and for recessive disorders. The major challenge still is to reliably distinguish between genetic information from the foetus and from the pregnant woman 5.

Inheritable diseases

Since 1999, the European Society of Human Reproduction and Embryology (ESHRE) has been collecting and analysing data on IVF use. The top 10 monogenic diseases for which PGD is used are: autosomal recessive diseases (β‐thalassaemia, cystic fibrosis, spinal muscular atrophy, sickle cell disease); autosomal dominant diseases [Huntington's disease (HD), myotonic dystrophy, Charcot–Marie–Tooth disease]; and sex‐linked diseases (fragile X syndrome, Duchenne muscular dystrophy, haemophilia). In several countries, an increasing number of tests are performed for adult‐onset diseases such as HD and cancer predispositions, in particular breast and ovary cancer. It seems that PGD is more acceptable to couples at risk than PND in order to prevent inheriting their mutations.

Overall, IVF/PGD is a safe and efficient treatment. The percentage of transferable embryos after PGD remains slightly higher than 40%, and highest in case of an autosomal recessive disease, followed by X‐linked and autosomal dominant disorders. A normal percentage of pregnancies ends in a miscarriage, and the number of terminations is extremely low. Using a definition of major malformation from the literature—that is, malformations that generally cause functional impairment or require surgical correction—the rate of abnormalities after PGD is not significantly different from ICSI children born without PGD. Embryo biopsy therefore does not add additional risk factors to the health of single children.

The example of Huntington's disease

Huntington's disease is a severe progressive neuropsychiatric late‐onset genetic disorder characterised by involuntary movements (chorea), cognitive deterioration and affective symptoms. There is no cure, and the only treatment is symptom relief and support. The mean age of onset is about 40 years; symptoms progress slowly, and death occurs after about 15 years. In 1993, the HD gene was isolated and shown that more than 40 triplet repeats of CAG disturb the normal function of the gene, which codes for a protein called huntingtin.

Presymptomatic or predictive testing for HD has been available since 1987 using DNA linkage, and since 1993 by direct mutation detection. However, PND of HD is still controversial. Some argue that selective abortion is unacceptable since a child born carrying the HD mutation might still expect many years of disease‐free life. Moreover, the test result itself has implications for the parents who may have decided not to be tested.

… the number of indications that might be addressed by genome editing is limited and PGD offers a safer alternative in most cases.

An alternative to direct testing is exclusion testing, which allows a parent with a 50% risk of HD to have children with a low risk. It screens the grandparental chromosome 4, transmitted by the parent at risk, using polymorphic markers closely linked to the HD locus. If chromosome 4 is inherited from the grandparent with the disease, the foetus has a 50% risk of having HD—the same risk as the asymptomatic prospective patient. However, since half of the foetuses carrying the grandparental allele would be normal, abortion is problematic 6.

Preimplantation genetic diagnosis is therefore a safe alternative. In the Netherlands, of all couples who tested for HD, 132 underwent PND with 262 pregnancies, 54 started PGD, and 25 used both. 76.5% of couples who used PND had one or more unaffected children, and 44.4% of couples who used PGD had at least one child. Couples reconsidered their choices in every subsequent pregnancy based on their previous experience, personal beliefs and the gender of the at‐risk partner 7.

Mitochondrial diseases

Mitochondrial diseases are a group of inheritable disorders on the maternal side, which are caused by a defect in mitochondrial oxidative phosphorylation. To avoid the transmission of mutations in mitochondrial DNA (mtDNA), three options are available: PGD or PND, use of oocyte donation and mitochondrial replacement therapy.

Usually, mtDNA mutations are heteroplasmic, and symptoms arise only if the number of mutated mitochondria raises above a certain threshold. PGD and PND decrease the risk of a severely diseased child, but this approach has several drawbacks. For instance, there remains an uncertainty whether the mutation load in the tested sample is representative for the rest of the embryo/foetus. There may also be changes in mutation load during embryonic/foetal development, and it is difficult to establish a cut‐off value for embryo selection owing to the unclear correlation between mutation load and clinical symptoms. From a clinical point of view, there are questions about efficiency, because there is no guarantee that a woman with a low heteroplasmic load in her somatic tissues produces oocytes with a similar low mutation load for a successful PGD cycle 8.

These regulatory differences along with the availability of particular tests and financial reasons have driven many patients to seek treatment abroad.

One alternative is the use of donated oocytes. Oocyte donation guarantees that the child will be free of the mtDNA mutation. However, it will have no genetic relationship with the mother. Another alternative is mitochondrial replacement therapy to combine the maternal genes with the normal mtDNA from a donated oocyte by nuclear transfer and then fertilise the oocyte with the husband's sperm. The first baby using this approach was born in September 2016 to a mother with Leigh syndrome, a fatal disorder that affects the developing nervous system, which had caused the deaths of her first two children. It is to be expected that the birth of this “three‐parent baby” will be followed by others.

Genome editing

With the advent of TALEN, zinc finger nucleases and CRISPR/Cas, it is now possible to deliberately change the genome of gametes and embryos and thereby the germline of future generations. Yet, human germline modification has for many years been widely considered off‐limits, both for safety and social reasons, and it is formally prohibited in more than 40 countries. However, this might change: in February this year, a scientific advisory group by the US National Academy of Sciences and National Academy of Medicine supported the use of germline editing to prevent the transmission of severe, inheritable diseases.

In terms of feasibility, the CRISP/Cas9 system has made gene editing much simpler, cheaper and efficient and greatly improved somatic gene therapy and genome editing 9. Editing the embryo genome still has major technical drawbacks, such as incomplete editing, leading to mosaic embryos, and off‐target effects. As newer, more accurate and therefore safer CRISPR/Cas9 systems are being developed, these technical limitations will likely be resolved. Nevertheless, the number of indications that might be addressed by genome editing is limited, and PGD offers a safer alternative in most cases. Only if one partner is homozygous for an autosomal dominant disease, or if both partners suffer from the same autosomal recessive disease, PGD cannot be used to select healthy embryos and genome editing might be an alternative. Less stringent indications could include cases where the number of embryos after PGD is small, when couples have a risk of transmitting more than one genetic disease, or when HLA‐matching embryos are selected to cure a sibling affected by a monogenic disease.

Regulatory framework and availability of PGD in Europe

There are different legal and regulatory frameworks for ART across Europe with direct consequences on clinical practice. Some countries allow PGD in a regulated environment, either by law as in the UK and in Belgium, or by a licencing system as in the Netherlands. Germany recently changed its law to legalise PGD, but it still limits access to other tissues than polar body biopsies for genetic analysis and its public healthcare system does not reimburse patients.

Most European countries have also set more or less strict limits for PGD indications. Social sexing—selecting the sex of the child for social reasons—is not allowed in almost all European countries. HLA typing without PGD for a genetic disease is not allowed in a number of countries; saviour sibling is only allowed if PGD is needed anyway. In regard to the indications, some countries allow PGD and PND for all conditions, while others have again set limits. Fatal childhood diseases are almost always accepted indications. The major differences comprise serious chronic diseases and adult‐onset diseases.

… we will also likely see an expansion of PND as non‐invasive diagnostic methods based on analysis of foetal DNA in the maternal blood become available.

These regulatory differences along with the availability of particular tests and financial reasons have driven many patients to seek treatment abroad. However, there are certain disadvantages to such cross‐border movements. If patients are not referred properly, they are left to identify/search for clinics themselves, potentially depriving them of the benefit of medical advice, counselling and support in a very vulnerable and emotionally challenging situation. Second, even if patients receive treatment abroad, the prohibition of PGD in their country of origin may complicate monitoring and follow‐up after treatment. Clinics could also be reluctant to get involved in follow‐up care of children born as a result of a prohibited treatment. Third, if PGD is prohibited, only more affluent patients can afford expensive treatment abroad.

The legal framework in the Netherlands as an example

In January 2003, the Ministry of Public Health, Welfare and Sport adopted a new licencing system that allows genetic centres to perform prenatal and postnatal diagnosis under the Dutch Special Medical Procedures Act (Wet op de Bijzondere Medische Verrichtingen). On top of the existing licences for the eight Dutch Clinical Genetic Centres, the Maastricht centre was granted a special permit to perform PGD. The indications did not change until 2008, when the Deputy Minister of Public Health, Welfare and Sport tried to expand the use of PGD by explicitly allowing the diagnosis of BRCA1 and BRCA2 mutations. However, the proposal was blocked by the smallest coalition party and a governmental crisis ensued. The controversy centred around the incomplete penetrance, variable expression and the therapeutic options available for hereditary breast and ovary cancer.

After 1 month of discussions and negotiations, an overwhelming majority of Dutch MPs voted in favour of a compromise proposal on testing embryos for hereditary diseases. The government allowed the Maastricht University Medical Centre, the only facility in the Netherlands where embryo selection is carried out, to test for a wider range of diseases. As part of the compromise, the government also decided to create a National Indications Commission to determine for which new diseases embryo testing can be carried out, provided that a number of criteria are taken into account: severity and type of the disease; options for prevention and treatment; medical criteria such as maternal age, premature ovarian failure or excessive BMI; and psychological burden for patients and ethical factors.

Reimbursement issues

The European Society of Human Reproduction and Embryology conducted an EU‐wide survey in 2008 about the reimbursement situation for PGD treatments 10. At that time, PGD was not allowed in four of 27 EU member states (LV, LT, DE, AT). Ten of 23 EU member states fully reimbursed PGD (FI, SE, DK, UK, BE, FR, ES, IT, SI, CY), five of which (ES, FI, SE, DK, UK) only reimburse PGD treatment in public clinics. Ireland reported that PGD was not practiced, although it is legal. In Spain, PGD was reimbursed when performed in public clinics, but most treatments took place in private clinics. Since a number of changes have taken place, a comprehensive overview is not available. In the Netherlands, for instance, public health care has been reimbursing three cycles of PGD per couple since 2006. If during one of these cycles a pregnancy is achieved, a new series of three cycles will be reimbursed again. This policy has increased the number of PGD treatments almost 10‐fold from about 60 cycles in 2006 to about 500 in 2016.

Both germline editing and PGD require some form of ART, which will be burdensome, emotionally stressing and a costly medical treatment…

Couples who seek ART because at least one of them carries an inherited mutation for a severe risk already have a huge emotional burden fearing that their child could be born with a devastating disease. They need not just medical assistance but also counselling and advice to help them cope with their situation and to find the right option to fulfil their desire to have children. In most cases, they opt for PND and/or PGD. Nonetheless, as there is no guarantee for success and considerable emotional stress during the treatment, it will be very unlikely that PGD will ever be widely used unless to prevent the transmission of a devastating disease. Financial issues also play a role as ART and PND/PGD incur considerable costs; in countries where the treatment is not covered by public health care, this leaves this option mainly for affluent couples.

In the coming years, we will also likely see an expansion of PND as non‐invasive diagnostic methods based on analysis of foetal DNA in the maternal blood become available. At the same time, PGD is already increasingly used to diagnose and prevent adult‐onset autosomal dominant diseases—such as breast cancer or HD. Moreover, gene editing technologies such as CRISPR/Cas will become safe enough to allow germline genome editing in the future. Even if their use to edit the human germline is currently prohibited, there is a clear demand albeit only for a limited number of diseases since PGD offers a safer alternative. Both germline editing and PGD require some form of ART, which will be burdensome, emotionally stressing and a costly medical treatment for couples with a serious medical condition. Therefore, it is unlikely that it will easily expand to offer services to healthy couples, such as enhancement, as some critics fear.

Conflict of interest

The author declares that he has no conflict of interest.