Choosing between possible lives

Oxford Journal of Legal Studies, Vol. 26, No. 1 (2006), pp. 153–178 doi:10.1093/ojls/gqi048 © The Author 2006. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org Choosing between Possible Lives: Legal and Ethical Issues in Preimplantation Genetic Diagnosis ROSAMUND SCOTT* Abstract—This article critically appraises the current legal scope of the principal applications of preimplantation genetic diagnosis (PGD). This relatively new tech- nique, which is available to some parents undergoing in vitro fertilization (IVF) treatment, aims to ensure that a child is not born with a seemingly undesirable genetic condition. The question addressed here is whether there should be serious reasons to test for genetic conditions in embryos in order to be able to select between them. The Human Fertilisation and Embryology Authority and the Human Genetics Commission have decided that there should be such reasons by broadly aligning the criteria for PGD with those for selective abortion. This stance is criti- cally explored, as are its implications for the possible use of PGD to select either against or for marginal features or for significant traits. The government is currently reviewing the legal scope and regulation of PGD. 1. Introduction The unfortunate phrase ‘designer baby’ has become familiar to debates surround- ing assisted reproduction. As a reference to current possibilities the term is mis- leading, but the phrase indicates a societal concern of some kind about control and choice in the course of reproduction. This article focuses on the relatively new technique of preimplantation genetic diagnosis (PGD), which is a way of seeking to ensure that a child is not born with an apparently undesirable genetic condition. It is an option available to some couples undergoing in vitro fertiliza- tion (IVF) treatment. Although success cannot be guaranteed, the purpose of the technique is to enable parents to have a child without a genetic impairment and, in so doing, to avoid moral or other difficulties in the termination of an already- begun pregnancy. People can choose to reproduce. They also have an interest in choosing, to some extent, the conditions under which they do so.1 However, it is *School of Law and Centre of Medical Law and Ethics, King’s College London. I am very grateful to Roger Crisp, Jonathan Glover, John Stanton-Ife and Steve Wilkinson for thoughtful comments on earlier drafts of this paper and to Bobbie Farsides, Mike Parker and Riccardo Rebonato for discussions relating to it. 1 For scholars supporting this view, see below nn 52 and 53. at Zhejiang University on December 9, 2011 http://ojls.oxfordjournals.org/ D ow nloaded from 154 Oxford Journal of Legal Studies VOL. 26 always difficult and controversial to delineate the moral and legal scope of this interest. In this context there is concern about the degree to which we should permit ‘negative’ testing and selection, that is, selection against impairments at the embryonic stage. Currently this is closely regulated such that the criteria for PGD are broadly in line with those for selective abortion; for this reason, nega- tive selection against marginal features (if technically possible) would not be per- mitted. There is also concern about the possibility of ‘positive’ selection, that is, selection in favour of features, be they marginal or significant.2 The question explored here is whether the interests and values at stake in PGD require a non- trivial justification for the decision to test for certain conditions or features and then to choose one or two (rarely three) embryos, resulting in the demise of those not chosen.3 The appropriate legal scope of PGD was the subject of a consultation exercise (in the form of a ‘Consultation Document’),4 after which the Human Fertilisa- tion and Embryology Authority (HFEA) and the Human Genetics Commission (HGC) produced a report, which I shall refer to as the ‘Outcome Document’.5 2 As J. Glover notes, the distinction between ‘negative’ and ‘positive’ selection ‘is sometimes a blurred one, but often we can at least roughly see where it should be drawn’. J. Glover, What Sort of People Should There Be? (Harmondsworth: Penguin, 1984) 31–2. A limited type of positive selection is currently in use and was considered in The Queen on the Application of Quintavalle v Human Fertilisation and Embryology Authority [2003] EWCA Civ 667 (CA), [2005] 2 All ER 555 (HL). The issue in this case was whether the HFEA could license HLA typing by PGD to enable a child to be born who would be a tissue match for a sibling with beta thalassaemia major. In the Court of Appeal ([2003] EWCA Civ 667), Phillips, Schiemann and Mance LJJ all held that PGD for the purpose, not sim- ply of ensuring that a woman was able to bring a pregnancy to term (as Maurice Kay J had held below), but also to ensure that a child would be free from genetic abnormalities, is lawful under the HFE Act 1990. The key statutory provisions requiring interpretation are s 2(1) and para 1(1)(d) of Sched 2. Section 2(1) states: ‘(1) In this Act. . . “treatment services” means medical, surgical or obstetric services provided to the public or a section of the public for the purpose of assisting women to carry children’. Para 1 of Sched 2 provides: ‘(1) A license under this para- graph may authorise any of the following in the course of providing treatment services – . . . (d) practices designed to secure that embryos are in a suitable condition to be placed in a woman or to determine whether embryos are suitable for that purpose . . . (3) A license under this paragraph cannot authorise any activity unless it appears to the authority to be necessary or desirable for the purpose of providing treatment services’. In July 2004 the HFEA announced that selection of an embryo which would have the same tissue type as an existing ill sibling would be legal even where the embryo itself was not at risk of a serious genetic condition. Permission must still be sought from the HFEA in each case and the procedure should be a ‘last resort’. ‘HFEA Agrees to Extend Policy on Tissue Typing’, 21 July 2004. The earlier decision on the case of the Whittaker family was contrary to this position. In April 2005, the House of Lords ([2005] 2 All ER 555), unanimously dismissed the appeal against the Court of Appeal judgment, which also questioned the HFEA’s revised policy. Lord Brown defined the issue raised by the case as ‘whether the [HFEA] . . . is empowered by the 1990 Act to license tissue typing’. (At para 42.) He went on to state that the ‘[t]he critical question . . . is whether tissue testing is a practice designed to determine whether an embryo is suitable for placing in a woman (para 1(1)(d)) and necessary or desirable for the purpose of providing a medical service which itself is to assist a woman to carry the child (section 2(1))’. (At para 49.) The House held, per Lord Hoffman (at para 35), ‘that both PGD and HLA typing could lawfully be authorised by the authority as activ- ities to determine the suitability of the embryo for implantation within the meaning of paragraph 1(1)(d)’. For dis- cussion of the issue of HLA typing in this context see S. Sheldon and S. Wilkinson, ‘Hashmi and Whittaker: An Unjustifiable and Misguided Distinction?’ (2004) 12 Med Law Rev 137. Although I make further reference to the Quintavalle case, this article concentrates on PGD in the general (i.e. non-HLA typing) case. 3 HFEA Press Release, ‘HFEA Reduces Maximum Number of Embryos Transferred in Single IVF treatment from Three to Two’, 8 August 2001. The HFEA’s commitment to reducing the number of multiple births has been strengthened in its Code of Practice—Sixth Edition, which only allows for the implantation of three embryos in women over 40. Embryos which are not chosen will be discarded or used in research. Freezing may also be a pos- sibility, but then the intention would be to implant some of the embryos in the future. 4 HFEA and Advisory Committee on Genetic Testing (AGCT), Consultation Document on Preimplantation Genetic Diagnosis, November 1999. 5 HFEA and HGC, Outcome of the Public Consultation on Preimplantation Genetic Diagnosis, June 2001. at Zhejiang University on December 9, 2011 http://ojls.oxfordjournals.org/ D ow nloaded from SPRING 2006 Choosing between Possible Lives 155 This followed the ‘precautionary principle’6 of broadly aligning the criteria for PGD with those of section 1(1)(d) of the Abortion Act 1967 (as amended by the Human Fertilisation and Embryology (HFE) Act 1990). That section requires that, in authorizing an abortion, two doctors judge there to be a ‘substantial risk’ of a ‘serious handicap’ in the child that would be born.7 The Outcome Document also suggested certain factors of further relevance to PGD decisions. In this way, just as serious reasons are legally required for selective abortion, so are they required for PGD. This stance is critically explored here. I argue that PGD to select against serious disabilities is justifiable. By contrast, and if it were possible, PGD for lesser, for instance aesthetic, reasons may be problematic under certain conditions. I explore the meaning of ‘seriousness’ in decision-making in this context and argue for the value of decisions made in this way. I also consider the (theoretical) use of PGD to select positively for meaningful traits such as intelligence, which raises different moral concerns. Although broadly appropriate, the alignment of the criteria for PGD with those for selective abortion results in some anomalies. The discussion raises questions about the suitable extent of parental input into decisions regarding PGD or selective abortion itself. Legally, questions about the appropriateness of a regulatory system which does not grant legal rights underlie aspects of the discussion. Morally, questions about the reasons for which decisions could be made in the reproductive context and so about the value—or lack—of reproductive autonomy are at stake. On an introductory theoretical note, the conception of autonomy and rights on which this analysis hinges combines aspects of the Interest and Choice theories of rights.8 This is because of limitations in either one approach. On the one hand, an interest-based account of rights is better placed to explain the signifi- cance of moral rights, since these can then be seen to protect important interests rather than just choice per se. Indeed, H.L.A. Hart has acknowledged that his 6 Ibid, para 25. On the possible interpretations of such an approach, see the Government Response to the Report from the House of Commons Science and Technology Committee: Human Reproductive Technologies and the Law (2005) paras 5 and 6, in response to the Committee’s Rec 2, para 46 and Rec 3, para 47, below n 40. 7 The relevant part of s 1(1) of the Abortion Act 1967 (as amended by the HFE Act 1990) reads: ‘Subject to the provisions of this section, a person shall not be guilty of an offence under the law relating to abortion when a preg- nancy is terminated by a registered medical practitioner if two registered medical practitioners are of the opinion, formed in good faith . . . (d) that there is a substantial risk that if the child were born it would suffer from such physical or mental abnormalities as to be seriously handicapped’. Whether the term ‘handicap’, ‘impairment’ or ‘disability’ is used in this context is not a neutral point. Exclusive use of the term ‘disability’ is associated with a purely ‘medical’ model of disability which has been discredited, though not successfully replaced, by the ‘social model’. Proponents of either model have accepted to some degree the limitations of their respective approaches and, as Jonathan Glover has recently noted, ‘[i]t is time to give up this debate, as it is now unfruitful for the same reason that makes the “nature-nurture” debate unfruitful. To many disabilities, there is a contribution from a vari- ety of sources, including functional limitation and social context’. J. Glover, Choosing Children—Genes, Disability and Design (Oxford: OUP, 2006) 7–8 (footnote omitted). Since some combination of the ‘medical’ and ‘social’ models is best placed to account for the significance of impairment or disability, I interchange the terms ‘disability’ and ‘impairment’. 8 On the Interest theory, for instance, Joseph Raz argues that to say that an individual has a right means, ‘other things being equal, an aspect of x’s well-being (his interest) is a sufficient reason for holding some other person(s) to be under a duty’. J. Raz, ‘Right-based Moralities’ (1982) in J. Waldron (ed.), Theories of Rights (Oxford: OUP, 1984) 182, 183. The predominantly countervailing Choice Theory is H.L.A. Hart’s, as in his ‘Are There Any Nat- ural Rights?’, in J. Waldron (ed.), Theories of Rights (Oxford: OUP, 1984) 77. at Zhejiang University on December 9, 2011 http://ojls.oxfordjournals.org/ D ow nloaded from 156 Oxford Journal of Legal Studies VOL. 26 Choice Theory cannot adequately explain either all legal rights or, more particu- larly, those which are part of social and political morality.9 On the other hand, since a purely Interest-based account emphasizes the idea of passive benefit, it is hard-pressed to explain, for instance, a right such as that to refuse unquestiona- bly beneficial medical treatment. As the Interest approach became more refined, the distinction between the active and the passive was replicated within the Interest Theory itself so that, on some approaches, it is particularly those inter- ests associated with autonomy which are sufficiently important to ground the ascription of rights.10 (The interests at work in purely Interest-based accounts are non-autonomy-related interests.) An implication of the ‘hybrid’ account adopted here is that autonomy need not itself be rights-based. Rather, as Joseph Raz argues, it is a value to which cer- tain interests may be related and from which certain rights derived.11 For instance, autonomy is a prominent value in the medical treatment context; it spawns the interests in self-determination and bodily integrity, which in turn underlie the right to consent to and refuse medical treatment. With regard to abortion and PGD, an important value is reproductive autonomy; it spawns an interest in avoiding reproduction as well as a sharply contested interest—which is the focus of discussion here—in determining, by PGD or (to a lesser degree) selective abortion, the kind of child who may be born. This interest is controversial in scope and so, both morally and legally (where a legal right is recognized), is the right to choose which child will be born. On this approach to the link between autonomy, interests and rights, repro- ductive autonomy is most valuable when there are good or serious reasons for certain choices. These imply that the moral interests derived from it are seriously invoked. 2. The Scope of Current Possibilities and Background Legal Framework It is appropriate to outline the current scope of PGD. The HFEA describes the technique as ‘typically involv[ing] several stages: the creation of an embryo 9 H.L.A. Hart, ‘Bentham on Legal Rights’ in A.W.B. Singer (ed.), Oxford Essays in Jurisprudence, Second Series (Oxford: OUP, 1973) 196–8. 10 See J. Waldron (ed.), ‘Introduction’, above n 8. For further discussion of the basis of rights particularly in the medical context, see R. Scott, Rights, Duties and the Body: Law and Ethics of the Maternal-Fetal Conflict (Oxford: Hart Publishing, 2002) 15–20. Generally see also A. Buchanan, D. Brock, N. Daniels, D. Wikler, From Chance to Choice: Genetics and Justice (Cambridge: Cambridge University Press, 2000) 207 for the view that rights arguments need support, which can be found by showing how rights protect important underlying interests and choices. 11 (Simplifying a complex argument) Raz suggests that given that society at large is involved in facilitating the existence of the conditions which enables the pursuit of an autonomous life, it is unlikely that the interest of any one individual can justify imposing such an onerous duty on so many. It follows that there can be no right to per- sonal autonomy as such, which means personal autonomy is one example of an ideal or value at the heart of our morality. J. Raz, above n 8, esp. 186–95. On Raz’s view, derivative rights protect and advance aspects of the indi- vidual’s autonomy and contribute to making autonomy possible. Ibid at 195. at Zhejiang University on December 9, 2011 http://ojls.oxfordjournals.org/ D ow nloaded from SPRING 2006 Choosing between Possible Lives 157 in vitro, the removal of one or more blastomeres, the genetic testing of those blastomeres for certain conditions and the transfer of suitable embryos to a woman’.12 The blastomere(s) (cells) are removed when the embryo consists of six to ten cells. There are three main categories of disease for which PGD is helpful:13 sex-linked disorders such as Tay-Sachs14 or Duchenne muscular dystrophy;15 single gene defects, for instance cystic fibrosis; and chromosomal disorders such as translocations, inversions and chromosome deletions. The related technique of preimplantation genetic screening (PGS) focuses on aneu- ploid embryos. These may be inherently non-viable (so that they either fail to implant or miscarry) and so PGS is particularly, though controversially, linked with attempts to increase the success rate of IVF.16 Aneuploid embryos may also develop into children with a condition such as Down’s syndrome. The technique of PGS was approved some time after the Outcome Document.17 ‘It is expected that . . . [it] . . . will only . . . be used’ in relation to:18 (i) Women over 35 years of age (ii) Women with a history of recurrent miscarriage not caused by translocations or other chromosomal rearrangements (iii) Women with sev- eral previous failed IVF attempts where embryos have been transferred or (iv) Women with a family history of aneuploidy not caused by translocations or other chromosomal rearrangements. 12 HFEA, Code of Practice—Sixth Edition, above n 3, para 14.1. 13 F. Flinter, ‘Preimplantation Genetic Diagnosis Needs to be Tightly Regulated’ (2001) 322 BMJ 1008, 1008–9. 14 ‘The classical form of Tay-Sachs disease (TSD) is a fatal genetic disorder in children that causes progressive destruction of the central nervous system . . . By about two years of age, most children experience recurrent sei- zures and diminishing mental function. The infant gradually regresses, losing skills one by one, and is eventually unable to crawl, turn over, sit, or reach out. Other symptoms include increasing loss of coordination, progressive inability to swallow and breathing difficulties. Eventually, the child becomes blind, mentally retarded, paralyzed, and non-responsive to his or her environment. To date, there is no cure or effective treatment for TSD’. National Tay-Sachs and Allied Diseases Association, Inc. http://www.ntsad.org/pages/t-sachs.htm. 15 According to the Muscular Dystrophy Campaign, ‘[t]his is a very serious condition. Most affected boys develop the first signs of difficulty in walking at the age of 1 to 3 years and are usually unable to run or jump . . . they often struggle to climb stairs . . . Rising from the floor can also prove difficult. As the condition progresses . . . [they] . . . are unable to walk as far or as fast as other children and may occasionally fall down. Some . . . also have learning and or behavioural difficulties, which may begin to manifest at this stage. By about 8 to 11 years . . . boys become unable to walk and by their late teens or twenties the condition is severe enough to shorten life expect- ancy. There are however many forms of management which are now available, which have changed t