The welfare of the child principle and the use of PGD: selecting for disability

5   The welfare of the child principle and the use of PGD: selecting for disability

Sarah Elliston

Introduction

Campbell and Cabrera argue in this volume ‘that prospective parents should be free to take such steps as they think fit to have what they regard as healthy and capable children’, adding the caveat ‘provided this does not cause undeniable harm, to these children or to other people, which is sufficiently serious to outweigh the presumptive case for reproductive freedom’.¹ Notions of parental freedom and harms to children will be explored in a more specific context in this chapter: that of preventing the use of embryos for treatment based on their genetic make-up.

The enactment of the Human Fertilisation and Embryology Act 2008 (HFEA 2008) may be viewed as enhancing the scope for choice of those who wish to become parents through assisted reproduction in the UK, largely due to its provisions concerning the possibility of founding families outside the two parent heterosexual model and its minor revision of the welfare principle.² The approach taken by the HFEA 2008 was to a considerable extent based on the House of Commons Select Committee on Science and Technology (HCSTC) report: Human Reproductive Technologies and the Law.³ This report had been criticised by some members of the Committee for having taken an ‘extreme libertarian approach’, their dissenting views being reflected in a special report issued at the same time as the substantive report.⁴ However, in one area the HFEA 2008 has taken a uniquely restrictive step. For the first time in the UK, legislation has sought to make it impermissible, at least in certain circumstances, to choose to try to have children that are known to have particular genetic constitutions. Testing of gamete donors may enable these kinds of decisions to be made before embryos are created and restrictions are imposed on preferential selection of donors where there is a significant risk of genetic abnormalities being passed on to children.⁵ However, the principal technique under consideration here will be prenatal genetic diagnosis (PGD), involving genetic testing of embryos.⁶ The idea that a state would involve itself through legal regulation in decisions about the genetic makeup of children and intrude on the reproductive choices of potential parents in this way raises troubling questions, despite the apparently beneficent motive of preventing children being born with potentially devastating genetic conditions. While concerns about the appropriate weight to be given to the welfare of children born through assisted reproductive technology were clearly prominent during the reviews of existing legislation and practice that led to the passage of the HFEA 2008, there was also a perceived need for clarity concerning when techniques such as PGD should be deemed to be acceptable and legally permissible.⁷ It will be argued here that if achieving clarity was a principal aim of the new legislative provisions, they may in fact fall short of achieving this goal. It will be suggested that they may provide some scope for interpretation that would allow potential parents who wished to do so to attempt to select embryos for disability, despite the legislative intent to the contrary. Even so, the retention of the general welfare principle in the amended HFEA 1990 may still render such attempts null and void, highlighting the need to consider the appropriateness of a child welfare test in this context. It also raises the issue of whether the new provisions serve any practical purpose, if the end result could have been achieved through use of the welfare principle. It will be concluded that their real effect is most likely to be in the message they may be perceived to send about the undesirability of having children with genetic disorders.

The HFEA 2008 amends in many significant ways the original statute regulating particular methods of medically assisting reproduction: The Human Fertilisation and Embryology Act 1990 (HFEA 1990). Hereafter they will be referred to collectively as the HFE Acts,⁸ save when reference to a specific Act is necessary. The HFE Acts prohibit certain activities absolutely.⁹ Other specified activities can only be carried out with a licence issued by the Authority. The type of licence relevant to this discussion concerns activities in the course of providing treatment services.¹⁰ Performing such activities without a licence is a criminal offence carrying penalties of imprisonment up to ten years or a fine or both.¹¹

Schedule 2 of the amended HFEA 1990, specifies the kinds of activities that may be authorised under licence. Those critical to PGD are the following:

Similar restrictions are placed where testing involves gender-related genetic disorders.¹² These will not be discussed separately.

On an initial reading, these provisions may seem to allow for no discretion in the use of embryos found to have an abnormality or to carry a risk of illness or disability, once a condition has passed the requirement of sufficient seriousness, and this threshold must be passed to allow PGD to take place at all.¹³

In addition to these provisions, it is also important to note the general ‘child welfare’ principle which forms part of licence conditions. This must be considered in all treatment decisions governed by the HFE Acts:

13(5) HFEA 1990, as amended

A woman shall not be provided with treatment services unless account has been taken of the welfare of any child who may be born as a result of the treatment (including the need of that child for supportive parenting), and of any other child who may be affected by the birth.

A failure to comply with the requirements of s. 13 is not itself a criminal offence,¹⁴ unlike certain other practices prohibited by the HFE Acts, such as placing an embryo in a woman other than a permitted embryo.¹⁵ Nevertheless, although a breach of the standard licensing conditions in itself would not lead to prosecution, it could result in the HFEA taking proceedings to suspend or revoke a clinic’s licence, so they are very likely to be observed.¹⁶ Effectively, the licensing conditions act as a legislative restriction on the choices of couples seeking treatment services governed by the HFE Acts by limiting the freedom of providers to offer such services. The HFEA’s regulatory role has been reinforced by the introduction of a specific statutory duty upon it to promote compliance with the requirements of the HFE 1990 and with the Code of Practice it issues under section 25 of the HFE Acts.¹⁷ Its duties also now include maintaining a statement of the general principles which it considers should be followed in carrying out its activities under the Act and its functions in relation to such activities.¹⁸ It has been suggested that ‘the Authority has always defined its general role by reference to its statutory remit. This new requirement is likely to produce a more detailed breakdown of its role in respect of particular and specific activities and functions.’¹⁹

In order to examine the potential impact of the new s. 13 restrictions on treatment following embryo testing, it is useful to consider examples of how they might be applied. For the sake of simplicity, it will be assumed that both gamete providers are intending to parent any child born as a result of the assisted reproduction service.

The application of the new embryo-testing provisions

PGD results in a possible choice within a batch of embryos –
couple does not wish to use an affected embryo

In this example, the couple has a number of embryos in storage and is seeking to avoid a specific disorder which comes within the terms of being a condition that can be tested for using PGD. All of the embryos are tested, and the results show that only one of these embryos is affected. The couple chooses to avoid using the affected embryo.

This is the most likely scenario to arise in practice. The vast majority of people who seek to use PGD will do so precisely to avoid particular genetic disorders due to a family history of a condition or increased risk of disorders, such as those associated with conception of children by older gamete providers. Should such a condition be detected in a particular embryo following PGD, their conclusion will almost inevitably be to reject this embryo for implantation in favour of one that does not have this condition. In such situations, there is no need to bring into play any of the new section 13 restrictions, as the potential parents’ preference is in line with them.

However, matters may not be entirely simple even here. Schedule 2 of the HFEA 1990 as amended contains further provisions concerning activities which may be licensed, and added in embryo testing as one of these activities.²⁰ Embryo testing can only be carried out for one or more specified purposes. These include establishing parenthood and tissue-typing where a sibling suffers from a serious medical condition to enable selection of an embryo that could be a match for potential donation to the sibling. The purposes that are relevant here, however, are the provisions relating to testing for genetic abnormality.²¹

Schedule 2 S1ZA

(1)

(a)   establishing whether the embryo has a gene, chromosome or mitochondrion abnormality that may affect its capacity to result in a live birth,

(b)   in a case where there is a particular risk that the embryo may have any gene, chromosome or mitochondrion abnormality, establishing whether it has that abnormality or any other gene, chromosome or mitochondrion abnormality.

In addition, Schedule 2 1ZA (2) provides that a licence cannot be issued for embryo testing unless the Authority is satisfied ‘that there is a significant risk that a person with the abnormality will have or develop a serious physical or mental disability, a serious illness or any other serious medical condition’. Difficulties in the interpretation and application of condition-based criteria are discussed by Jackson in this volume.²²

Nevertheless, once a condition has been authorised by the HFEA as one that can be tested for, it would seem to follow that a decision can be legitimately taken to choose not to implant any tested embryo. Of course it is the case that a woman could not be forced to accept the implantation of any embryo due to the extreme violation of respect for her autonomy and bodily integrity that this would entail, and either gamete provider could withdraw consent to the use of embryos created with their gametes before treatment was provided.²³ However, the point is that the couple could prefer to use unaffected embryos to those that are known to have a genetic condition disclosed by PGD. That choice is consistent with the approach taken by the new s. 13 provisions which seek to avoid such embryos being used for treatment.

Of course, the obvious question might be why a couple would wish to use an affected embryo when they have others that could be used instead? Although it might be supposed that most potential parents would seek to avoid passing on genetic disorders, there are occasional examples of people seeking to make such reproductive choices. The scenario that has caused most discussion is where a couple deliberately seeks to have a child with a genetic condition for social reasons. The classic example is congenital deafness, illustrated by a deaf lesbian couple in the United States of America, although they did not actually use PGD. Ms Duchesneau and Ms McCullough sought to increase the chance of having a child who would be deaf by using a deaf friend as a sperm donor. Their decision has provoked considerable academic debate on the merits or otherwise of ‘choosing for disability’.²⁴ The couple contended that deafness is a culture, not a disability, and that rather than deafness reducing the child’s capacity, being able to hear would constitute a lack of capability in their child since a hearing child would be prevented from full participation in deaf culture.²⁵ Using PGD to select an affected embryo could make the difference between increasing the chances of having a deaf child to making this a certainty. Another example can be found in the condition of achondroplasia. Its most obvious feature is that it is the commonest cause of disproportionately short stature, or dwarfism, although it may be associated with other medical complications.²⁶ Achondroplasia is an FGFR3 gene mutation which shows autosomal dominance in inheritance. Where both partners carry the mutation, in natural conception there is a 25 per cent chance of an unaffected pregnancy, a 50 per cent chance of an affected pregnancy, and a 25 per cent chance of a pregnancy where the child carries two copies of the mutation, which is usually fatal shortly after birth. It has been suggested that a couple with achondroplasia might seek not only to avoid the possibility of conceiving a child with the lethal combination of genes but to have a child with achondroplasia, rather than one without.²⁷ Again the reasons for this would be parental preference for a child like themselves, or practical issues in caring for the child.²⁸

Although the circumstances in which there might be a deliberate attempt to use PGD to conceive a child with a particular genetic disorder are likely to be rare in practice, there is evidence that they have been made. Baruch et al. surveyed American clinics offering PGD and found that 3 per cent of respondents reported having intentionally used PGD ‘to select an embryo for the presence of a disability’.²⁹ It is exactly this kind of choice that s. 13 of the amended HFEA 1990 is directed against. Indeed this is made crystal clear in the explanatory note that accompanied a draft of the Human Tissues and Embryos Bill published in 2007. Following a description of the intention to make it a condition of treatment licences that embryos tested and known to have specified abnormalities were not to be preferred in treatment services to those which did not, and that this prohibition would also extend to the selection of persons as gamete or embryo donors, it went on to say:

[t]here have been reported cases outside the UK involving the positive selection of deaf donors in order deliberately to result in a deaf child. The new section [as numbered] would prevent this.³⁰

It follows that, in this scenario, a deliberate choice to select an affected embryo in preference to others that were unaffected would be prevented. This may pose particular problems for couples with conditions such as achondroplasia, where one possibility is that an embryo will have a combination of genes incompatible with survival. Although amniocentesis or chorionic villus sampling can be performed during pregnancy to detect double dominant mutations, the use of PGD would enable decisions to be made without having to consider abortion where this variation is found to exist. However, there may be couples with achondroplasia who, while they might not prefer to have a child who also has a single copy of the mutation and who is affected by the condition, would be content with this. The same of course may be true for other genetic conditions, but unless they have variants which are likely to be fatal or cause additional problems for the child, couples who have no preference would be unlikely to opt for PGD.³¹ If a woman had undergone screening during pregnancy, there would be no question of forcing her to undergo a termination of pregnancy, whether the foetus had a single or even double copy of the mutated gene. However, if PGD were undertaken to avoid the fatal mutation, then could a decision be made to use an embryo with the single mutation? According to the new provisions, it would appear not, at least if they had other embryos that were free from the mutation.³² While wishing to avoid a fatal condition, and the consequent decision that might arise as to whether to abort a pregnancy, it appears that women using PGD who might have chosen to carry an embryo with achondroplasia could be denied that opportunity under the new provisions.

However, it may be argued that the legislation was not intended to achieve this outcome and that there are means of interpretation that would avoid it. One possible solution might be to focus on the wording of the legislative provisions, which are couched in terms of ‘preference’. It could be suggested that if a couple were willing to have either a child with achondroplasia or one without, that they are not in fact expressing a preference for a child with a genetic disorder. If this were accepted, then they could be permitted to use an embryo chosen at random from the embryos that do not have the fatal genetic combination, but which may include embryos with and without the non-fatal mutation. This would appear to satisfy the intention of the provisions to prohibit deliberately using affected embryos for treatment but allow some room for manoeuvre for potential parental choice to run the risk of a child having what is likely to be regarded as a disabling medical condition. The potential parents in such a situation are not ‘preferring’ an embryo with such a condition, nor seeking to select one on this basis; they are merely accepting the possibility.

There are some grounds to suggest that the provisions might bear such an interpretation since, despite the clear intention behind them, it does not necessarily follow that even where there is a known risk that an embryo may have a genetic abnormality or be of a sex giving rise to the risk of a resulting child having a serious physical or mental disability, serious illness or serious medical condition, it cannot be implanted. This is apparent from the guidance given in the HFEA’s 8th Code of Practice in relation to embryo testing.³³ There is no absolute prohibition on implanting such embryos. In addition, although the Code of Practice omits to refer to the preferential selection of embryos here, the wording of the statute does appear to make ‘preference’ a necessary criterion for the restrictions on embryo selection to apply. Furthermore, there is no requirement that couples submit their embryos to PGD. Rather, the expectation is the opposite: that PGD is currently seen as requiring justification in defined circumstances and is not to be used as routine practice. There is no statutory requirement that a couple with a known inheritable disorder, or where there is a clear risk of a genetic abnormality in an embryo, have to agree to PGD on embryos created with their gametes before they can use them in treatment. Similarly, donors who are known to have such disorders or be at risk for them are not prevented from providing gametes under s. 13 (9); it is only preferentially selecting them that is prevented. In fact, it is possible that limitations on providing treatment to couples or using donor gametes might be placed where there is a known risk of embryos having genetic disorders, but these would need to rest on a different provision – that of the general welfare of a resulting child, contained in s. 13 (5). This will be explored further in due course. For the moment, I will turn to an even more contentious possibility – that the prohibitions on embryo selection may not prevent a deliberate attempt to select for an embryo with a genetic disorder.

PGD undertaken on one embryo at a time – couple wishes to use an affected embryo

Here the couple chooses not to undertake PGD on a batch of embryos, but rather only one embryo at a time is created and tested. For the sake of argument, the first embryo tested has a genetic disorder that is contained in the HFEA’s authorised list of conditions for testing, but the couple wishes to use it for treatment.

My contention would be that the conditions set out in ss. 13 (9–11) may not prevent them from making this choice. This is because of the wording of these provisions which is expressed in terms of ‘preference’ in selection, comparing the affected embryo with others that are ‘not known’ to have the abnormality or carry the risk.

This is reinforced by the guidance given in the HFEA’s 8th Code of Practice on the interpretation of the restrictions on selecting embryos for treatment in s. 13 of the HFEA 1990, as amended.

However, this is not what the provisions say. They require a preference to be expressed in comparison with embryos that are not known to be affected. This still would seem to permit a wide or a narrow scope for such a comparison to be made. There are strong arguments against the idea of a very widely drawn category of comparator embryos. It may be contended that there should be a realistic possibility of the couple being able to access and use an alternative embryo. For example, the fact that other people have embryos stored for their own treatment that are not known to be affected by genetic disorder would not allow the couple with the affected embryo to seek to use them. It would therefore be absurd to suggest that in seeking to use their own embryo, they are ‘preferring’ to do so rather than use these other embryos. This is not a choice that is open to them. This supports immediately narrowing the category of comparator embryos to those that could be available to the couple to use, which leaves only embryos that are donated for other people’s treatment and other embryos of the couple.

The next issue is whether the other embryos must actually be in existence at the time that the comparison is being made or whether the possibility of future unaffected embryos that they could use can be brought into consideration. The response to this is less obvious, since there is no guidance as to the timeframe in which the comparison need be made. Again, however, it can plausibly be argued that the preference which the couple expresses should be directed to options that are currently available to them, not predicated on some, as yet uncertain, prospect of being able to secure donated embryos or successful creation of embryos from their own gametes that are not known to have disorders. This would suggest that the comparator group that the s. 13 provisions are concerned with are other embryos that the couple has available to use at the time when decisions about the affected embryo are being made. This seems to be implied from the HFEA guidance which refers to ‘no other embryos being suitable for transfer’, which suggests that the comparator embryos must be available for transfer at the time that the decision is being made.³⁴

If this is accepted, and it is fair to say that this remains a speculative approach, then it would seem to follow that if the couple do not have any other embryos in existence, they would not fall foul of the ss. 13 (9–11) of the treatment provisions by seeking to use the affected embryo. As noted earlier, these do not provide a blanket prohibition on the use of embryos that have been tested and found to have particular genetic conditions; only upon preferring to use them to those that are not known to have such conditions if there ‘is at least one other embryo suitable for transfer that is not known to have the characteristics’.³⁵ For a couple determined to seek to have a child with a particular genetic makeup, the route might accordingly be to ask to have embryos created and tested serially and to consent to the disposal of embryos found not to have the condition until one is found to have it. Consent could be withdrawn for an unaffected embryo’s continued storage; it could be donated for treatment of others, training purposes or for research. Any of these options would render it unavailable for a couple’s own treatment.³⁶ Embarking on this route would rest on the feasibility of having gametes available to use to create embryos in this way. It is likely to involve egg freezing which, as the HFEA notes, ‘is still a relatively new and experimental technique’ and has a low success rate in maintaining viable gametes.³⁷ Although additional methods such as vitrification are becoming available which appear to improve the chances of eggs surviving the freezing and thawing processes:

[r]ecords show that up to 31/12/2008, around 6388 eggs have been stored in the UK for the patient’s own use. To date around 88 embryos from stored eggs have been created. These embryos were transferred to women in around 32 cycles, which resulted in around 3 live births. These figures are for both eggs which have been stored using slow freezing and vitrification methods.³⁸

This would also require the willingness of clinics to participate in what would clearly be an attempt to circumvent the statutory provisions. Nevertheless, such circumvention would not appear to be unlawful nor to directly contravene current HFEA policy. The HFEA’s 8th Code of Practice, having allowed for the possibility that embryos known to have genetic abnormalities can be implanted, goes on to say:

10.17
The use of an embryo known to have an abnormality as described above [in 10C] should be subject to consideration of the welfare of any resulting child and should normally have approval from a clinical ethics committee.

10.18
If a centre decides that it is appropriate to provide treatment services to a woman using an embryo known to have an abnormality as described above, it should document the reason for the use of that embryo.

NOTE: An example of an embryo not suitable for transfer in this context is one that has no realistic prospect of resulting in a live birth.

Therefore, even if the specific statutory restrictions on selecting affected embryos can be avoided, it is entirely possible that such a practice would still be caught by the more general restrictions of the HFEA 1990, if the implantation of an affected embryo was thought to sufficiently risk any resulting child’s welfare. There is, therefore, a two-stage process in considering whether embryos can be implanted where PGD is involved: first, the embryo-testing provisions in ss. 13 (9–11), and, second; the welfare provisions in s. 13 (5).

The development of the techniques of PGD was based on apparently beneficent grounds: to improve the success rates for assisted reproduction by not commencing a pregnancy with an embryo that has a serious disorder that might result in miscarriage; to avoid a choice having to be made as to whether to abort a foetus with a serious disorder identified during pregnancy; and, underlying this, concerns about the welfare of children.³⁹ The development of the role of the HFEA in regulating PGD has been discussed elsewhere in this volume and so will not be repeated here.⁴⁰ However, questions about the use of PGD to select for disability have been asked in a number of public consultations and reviews. For example, in 1999 a public consultation exercise was carried out by a joint working party of the HFEA and the Advisory Committee on Genetic Testing.⁴¹ After outlining the potential choices that could be made to test embryos and act on the results, including seeking to use an affected embryo, it asked respondents ‘Can the principle of the Welfare of the Child ever be compatible with the decision to begin a pregnancy knowing that a child will be born with a genetic disorder?’ The responses were as follows:

47% acknowledged that there might be situations in which the replacement of an affected embryo would be compatible with the principle of the welfare of the child. 31% argued that this should be a decision left to the parents. 22% argued that starting such a pregnancy could never be compatible with this principle.⁴²

Although the number of responses was small (98), and hence the weight that can be placed on them is limited, they revealed a range of factors that were thought to be relevant:

The majority of these responses qualified their answers by stating that this might only be appropriate in certain circumstances, depending on severity of the condition, family situation and other similar considerations. In addition, most of the ‘yes’ responses concerned only the question of affected embryos where there were no unaffected ones for transfer, but rejected PGD as a method of choosing an affected embryo for social or other reasons.
    It should also be noted that many respondents were unfamiliar with the terminology of ‘welfare of the child’ and some respondents questioned its meaning, range or relevance.⁴³

It is interesting to see some doubts being expressed about the meaning of the welfare of the child principle and how it might apply in this situation.

On the specific issue of selecting for disability, the HCSTC in its report entitled Human Reproductive Technologies and the Law was ambivalent and did not reach a final conclusion on whether embryos with, as it termed them, ‘undesirable characteristics’ should be able to be used in treatment. On this issue it said:

We can imagine that many clinicians would baulk at the idea of selecting, for example, a deaf child using PGD, but we do not feel that the creation of a child with reduced life opportunities is sufficient grounds for regulatory intervention, else we might logically deny poor people IVF. Professor Tom Shakespeare told us that PGD should not be allowed to select out ‘minor or trivial’ conditions such as restricted growth or deafness. [Ev 363] On this basis, it is difficult to argue that they should not be selected rather than deselected. A more challenging but unlikely scenario would be the desire to select a child who would suffer obvious discomfort (rather than disadvantage), or worse. In this area there needs to be further debate.⁴⁴

It was therefore not at the behest of the HCSTC that ss. 13 (9–11) of the provisions on embryo selection were included in the legislation, nor was any suggestion of this raised in the government response to this report.⁴⁵ However, a further consultation exercise undertaken by the Department of Health appears to have provided the basis for their inclusion in subsequent legislative proposals. This consultation specifically asked whether there should be a prohibition on deliberately screening in, or selecting for, impairments or disabilities as opposed to screening out, or selecting embryos free from impairments or disabilities.⁴⁶ Responses to the consultation were published in March 2006. There was no attempt in this document to draw conclusions, and it was noted that on the issue of screening in or selecting for impairments the responses demonstrated a range of perspectives.⁴⁷ Nevertheless, it appears likely that the impetus for including legislative provisions on this subject came from the HFEA’s response, which sought parliamentary guidance. Following this consultation, proposals for legislation appeared in a White Paper issued in December 2006. Despite the lack of any clear consensus or recommendation for this approach in the prior consultative processes, it was stated that ‘[d]eliberately screening-in a disease or disorder will be prohibited’.⁴⁸ While the sections in which this issue was addressed changed at different stages of the drafting of the Bill and its passage through Parliament, the fundamental principle of prohibiting a preference for affected embryos to those not known to be affected remained consistent.⁴⁹

As well as seeking to improve success rates in terms of live births in assisted reproduction, and in order to try to avoid the issue of abortion arising, as it might if abnormalities were discovered only during pregnancy, the inclusion of these provisions seems, therefore, to be rooted in some notion of the concept of child welfare.

The welfare of the child principle and child law

The need to take account of the welfare of children born as a result of treatment services governed by the HFE Acts may seem at first to be an uncontroversial proposition.⁵⁰ Concern to protect children is at the heart of international declarations such as the United Nations Convention on the Rights of the Child, which is the most widely ratified international convention in existence.⁵¹ This Convention states in Article 3 that:

[i]n all actions concerning children, whether undertaken by public or private social welfare institutions, courts of law, administrative authorities or legislative bodies, the best interests of the child shall be a primary consideration. [my italics]

In the UK this approach may be seen to be reflected in the general legal principles that govern decisions about children. In England and Wales these are set out in the Children Act 1989, which provides that when hearing cases under the Act, the courts are required to have the welfare of the child as their paramount consideration. There are similar provisions under the equivalent statute in Scotland, the Children (Scotland) Act 1995. In England and Wales there is also the inherent jurisdiction which is usually exercised in wardship proceedings, although it may be used independently. Here, the principle is that judicial decisions must be made in the child’s best interests.⁵²