Natural conception 

Perhaps surprisingly, one of the most compelling analogies between this proposal and existing reproductive practice is the analogy with natural conception. If any reproductive practice is ethical, then presumably reproduction as a result of natural conception, pregnancy, and birth is ethical. Yet natural conception does not guarantee that those embryos that are conceived will come to term. A significant percentage of embryos, up to 33% according to some authorities, will be spontaneously aborted before pregnancy or at some stage of the pregnancy (Modvig et al., 1990). A willingness to conceive naturally therefore requires that the couple be prepared to sacrifice those embryos that may be spontaneously aborted in the course of the attempt to become pregnant for the sake of the life of the child that is eventually born (Harris, 2006, Harris, 2007).

In fact, even this description exaggerates the moral weight of the justification for the destruction of embryos that occurs naturally. While, in those circumstances in which a pregnancy is planned, the parents intend the birth of a child, they are unable to justify this with reference the child’s benefit: as the child does not exist at the time at which this decision is made, the child may neither be harmed nor benefitted. Instead, parents’ reasons for wanting a child necessarily refer to the desires of existing persons, to have a family, to experience the joys of parenthood, to express their love for each other, or to provide a companion for an existing child. While these may be admirable desires, they do not seem to have as much weight as the desire to save the life of an existing child. Moreover, of course, many pregnancies are not planned and result instead from contraceptive accident, risk taking, passion, intoxication, or ignorance. In such cases, embryos are created and consequently often destroyed (when they fail to implant or miscarriage occurs) for reasons which are at best morally trivial and are often reprehensible.

Both natural conception and the creation of saviour embryos require a willingness to sacrifice embryos to serve the desires of existing persons. The reasons for the creation and destruction of saviour embryos are prima facie more morally compelling than the reasons for the creation and destruction of embryos in natural conception. As natural conception is – presumably – ethical, this suggests that the creation of saviour embryos would also be ethical. Of course, the pursuit of natural conception only requires a willingness to risk the destruction of embryos whereas the destruction of embryos is required by the application of saviour embryos. However, the risk of the destruction of embryos that is involved in natural conception is converted into a virtual certainty in another reproductive technology – in-vitro fertilization.

In-vitro fertilization 

Because of the costs, discomforts, and risks involved in each cycle of IVF and because the rate of successful pregnancies per embryo conceived is still low, IVF laboratories will usually create multiple embryos. These embryos will then be screened according to the IVF technician’s estimation of how likely they are to lead to a successful pregnancy, with the ‘best’ embryos first to be implanted into the womb of the woman who wants to become a mother. Once a pregnancy is secured, the remaining embryos will normally be discarded.

IVF therefore requires a willingness to create multiple embryos knowing that most of them will not be implanted and will eventually be destroyed (Devolder, 2005a,Harris, 2006, Shannon and Cahill, 1988, Singer and Wells, 1984). Moreover, while the screening involved in IVF is usually thought of as ‘screening in’ for implantation, it might equally well be thought of as ‘screening out’ embryos that are thereby unlikely to be implanted. These embryos are thus effectively ‘selected for destruction’ for the sake of securing a pregnancy leading to the birth of a(nother) child. This process arguably further instrumentalizes the embryos that are destroyed in the process (Shannon and Cahill, 1988). Finally, it is also the case in IVF that, after a pregnancy has been secured, it can be said that each embryo that was not implanted was created for the sake of another child.

Like the creation of saviour embryos, IVF involves the destruction of embryos, the selection of embryos for destruction, and the creation of embryos in the knowledge that they are likely to be destroyed. IVF is dedicated to bringing a child into the world, whereas the proposal under discussion would be aimed at saving the life of an existing child. If IVF fails, the desires of the parents are frustrated but no other individual is harmed; on the other hand, without the creation of a saviour embryo, a child will die. Again, the justification for the way embryos are treated in the creation of saviour embryos seems significantly greater than that for the same treatment in IVF.

However, it remains true of each individual embryo that is created in IVF, that it is created with the intention of bringing a child into the world (FitzPatrick, 2003). This is not true of this paper’s proposal, which involves creating human embryos with no intention of allowing them the opportunity to flourish. Another technology that would also involve the creation of embryos for a purpose which requires their destruction, therapeutic cloning, is discussed below. First, however, a number of other reproductive technologies that have elements in common with this paper’s proposal are explored.

Preimplantation genetic diagnosis 

Preimplantation genetic diagnosis extends the screening involved in ordinary IVF to the genetics of the embryos created in order to increase the chances that a child will be born healthy. Again, PGD will typically involve screening out undesirable embryos rather than screening in desirable embryos – although this distinction is not always clear given that screening out undesirable traits will also be screening in desirable traits (Silver, 1999). PGD will also involve the destruction of those embryos that are not selected for implantation. Whereas the creation of surplus embryos might be said to be an unintended consequence of IVF, brought about by the low rates of implantation of embryos, the creation of multiple embryos, most of which will subsequently be destroyed, is an essential part of PGD, which aims to select one embryo from amongst many. Moreover, when the procedure is initiated there is the intention that particular (sorts of) embryos, if detected, will not be implanted and will therefore be destroyed. It is this feature of PGD that has been singled out for criticism by some activists within the disability community on the grounds that PGD necessarily involves the belief that it would be better if disabled people did not exist (Asch, 1988, Asch, 2000, Kaplan, 1993, Saxton, 1998, Wendell, 1996).

If PGD is ethical then presumably not only the destruction of embryos but also the creation of ‘excess’ embryos, and the selection of embryos for destruction are ethical, and thus the presence of these elements in the creation of saviour embryos should not rule it out. Nonetheless, like IVF, PGD only involves the destruction of embryos as a foreseen but unintended consequence of the pursuit of a healthy baby. The procedure under investigation would require the destruction of an embryo and, as such, this would be an intended consequence of the procedure. However, the widespread availability of another reproductive technology – abortion – suggests that such destruction may sometimes be ethical.

Abortion 

Abortion is legally available in many polities and widely practiced even in those polities where it is not legal. It is difficult to see how those who accept the moral permissibility of abortion in any circumstances could object to the destruction of embryos involved in the production of saviour embryos. Insofar as abortion would usually be procured for a reason, the embryo might be said to be destroyed for the sake of this reason. Yet the reasons for abortion usually fall well short of the need to save the life of an existing child that might justify the creation of saviour embryos.

Unfortunately, in the context of debates around the proper treatment of human embryos, abortion is likely to be just as controversial as the creation of saviour embryos and so arguments by analogy from the moral permissibility of abortion will have limited traction when it comes to convincing critics of the destruction of embryos involved in making saviour embryos that such destruction is warranted. However, the analogy with abortion is worth mentioning because a significant percentage of persons do believe that abortion may sometimes be justified and thus that the destruction of embryos may sometimes be justified. If the destruction of embryos is ever justified, it seems it would be justified in the scenario envisioned here.

Of course, the justification (or otherwise) of abortion is only relevant to the ethics of the destruction of embryos, whereas the procedure imagined here would also involve the creation of embryos for the sake of saving the life of another child. The next analogy to explore therefore is the analogy with saviour siblings.

Saviour siblings 

The creation of saviour siblings has been the topic of extensive ethical debate, with a number of authorities concluding that it is in fact ethical (Damewood, 2001,Devolder, 2005b, Fost, 2004, Ram, 2006, Robertson et al., 2002, Sheldon and Wilkinson, 2004). It is also legally permissible in a number of jurisdictions (Ram, 2006,Spriggs and Savulescu, 2002). The main difference between the creation of saviour embryos and the creation of saviour siblings is that the latter involves the creation of a human child and not just of an embryo. After an embryo with an appropriate tissue type has been selected using PGD, this embryo is then implanted into the womb of a woman with the intention of extracting some of the child’s tissues (usually bone marrow or umbilical cord blood) for the purpose of transplant into a terminally ill child (Verlinsky et al., 2001. Early examples of the pursuit of a matching tissue donor were relying on brute luck to ensure that the child born could serve as a donor for an existing child (Ram, 2006) but the use of PGD maximizes the chance that the child born will be an appropriate source of tissue and minimizes the chances that multiple children might need to be brought into the world in order to achieve this result. Because the creation of a saviour sibling will usually require the creation of multiple embryos for the sake of PGD, it will also involve the destruction of embryos and the creation of embryos knowing that most of them will be destroyed.

However, in contrast with the hypothetical creation of saviour embryos, all of the embryos created in the process of creating a saviour sibling are conceived with the intention of bringing a child into the world. Of course, this is not the only intention involved in the creation of saviour siblings, as the procedure is initiated with the intention of thereby saving the life of the already existing child. The presence of this ‘other’ intention has led critics of this procedure to argue that it involves ‘instrumentalizing’ the child that is created by bringing it into existence as a means to the end of saving another child’s life (King, 2006, McBride, 1990, Sutton, 2004). Defenders of saviour siblings have responded that the Kantian injunction against instrumentalizing human beings prohibits the treatment of others solely as a means to an end and that the creation of a saviour sibling does not involve this because the parents of the child that is born will inevitably also love this child for its own sake (Boyle and Savulescu, 2001, Devolder, 2005b, Sheldon and Wilkinson, 2004). The significance of this claim is in turn tendentious because, while it may be true that parents will love the saviour sibling, it is in many cases less clear that they would have had this child were it not for the desire to source tissue for transplant; consequently, in making the decision to conceive a child they may well have been doing so solely as a means to an end.

If the creation of saviour siblings is ethical, it must be the case that the creation of human embryos for the purpose of saving a life of an existing child is ethical – although this in itself does not settle the further question of the ethics of creating embryos with no intention of creating a living child. However, if the objection to the treatment of the beginnings of human life in the hypothetical creation of saviour embryos relates to the instrumentalization of human beings then it might be argued that it is in fact preferable to instrumentalize an embryo rather than a child. Any child born as a result of the need for a saviour sibling will grow up with the realization that they were conceived for the sake of making tissue for transplant; the circumstances of their conception may have psychological consequences (King, 2006, Sutton, 2004). This will not be the case with the creation of saviour embryos where, if instrumentalization occurs, it occurs without any consequences for any particular person. In this important regard, the creation of saviour embryos is arguably more ethical than the creation of saviour siblings.

Use of ‘surplus’ embryos post-IVF 

As noted above, IVF will typically involve the creation of multiple embryos but only the implantation of a small subset of these. The majority of couples undergoing IVF will therefore be left with a number of frozen embryos after they have succeeded in having a child. The question of what to do with these ‘surplus’ embryos has been one of the most controversial and vexed ethical issues surrounding IVF (de Lacey, 2007, Singer and Wells, 1984). In a number of jurisdictions, including Australia, the option is now open to couples to make these embryos available to scientists for use for research (Knowles, 2004, Research Involving Human Embryos Act, 2002). The argument for this practice is compelling: as these embryos are going to be destroyed anyway it seems preferable that their existence should contribute to the possibility of improving human wellbeing in the future (Devolder, 2005a, Harris, 2006, Savulescu, 2000).

In so far as the use of embryos for research purposes is ethical, their use for therapeutic purposes would also seem to be ethical given that the latter would, ex hypothesi, result in an immediate and concrete benefit to identifiable human beings whereas the former involves only the possibility of some future benefit. However, it remains the case that research on surplus embryos involves using embryos that already exist rather than creating them for this purpose.

Before leaving this analogy, though, it is worth noting that the existence of surplus IVF embryos opens up the possibility of deriving cell lines for therapy without needing to create saviour embryos. For instance, if the couple who had a sick child had already undergone IVF they might be able to find a tissue match with one of their existing surplus embryos. They could then consent to the destruction of this embryo for the purpose of deriving stem cells and then the appropriate cell lines from it to use to save the life of their sick child. Given that this embryo would otherwise be destroyed – and especially if the parents have the right to donate the embryo for destructive research – it is difficult to countenance any objection to this procedure. Perhaps slightly – but only slightly – more controversially, the parents of the child requiring a transplant could seek the help of other couples who had undergone IVF and who had surplus embryos to see whether a tissue match could be found with any existing embryo anywhere. Again, as long as the appropriate consent was secured from those responsible for the embryo, it seems as though deriving cell lines to save the life of an existing child, from an embryo that would otherwise be destroyed, would be ethical.

Saviour donated siblings 

If the use of surplus embryos for therapeutic purposes is ethical, the possibility arises that the parents of a child requiring a stem cell transplant might arrange to provide an embryo to another couple in need of a donor embryo for reproductive purposes and undergo IVF and PGD in order to do so, in the hope that any surplus embryos created could serve as a source of tissue for their existing child. They would therefore enter the IVF programme with the intent to conceive a child. Once the other couple had secured the birth of a child using one of the embryos that the parents of the sick child had created, the parents of the sick child could then consent to allow their remaining embryos to be used for ‘research’ into the derivation of stem cells. If this derivation is successful, the parents of the sick child could then hope that these tissues might be used in therapy to save the life of their child. In such a scenario, the child born as a result of the donation could then be said to be a saviour sibling to the sick child, although the cells used to save the life of the sick child would in fact been derived from another embryo.

Of course, if the parents can find another couple willing to have a child conceived of their gametes, they might equally well request that PGD be used to ensure that this child could serve as a tissue-matched donor to save the life of their child – in which case there would be the creation of a saviour sibling born to a couple who are not the genetic parents of the child. To generate the precise scenario envisioned here, it would therefore have to be the case either that the tissues required for transplant could not be sourced except from embryonic stem cells or that the embryo donors could not rely upon the birth parents to consent to allow the child to serve as a tissue donor. In any case, the interesting question about this scenario is not whether or not it would be the only or the best way to achieve the parents’ goals but whether or not each individual step in the process described would be ethical.

Note that in the above scenario, all the embryos conceived are created with the intention of bringing a child into the world. This procedure would therefore avoid objections based on the moral impermissibility of the deliberate creation of embryos for the purposes of destroying them. Note also that the couple seeking a donated embryo have a more plausible case that they will love the resulting child for its own sake than the parents of ordinary saviour siblings. The creation of saviour donated siblings in this scenario therefore appears to be more ethical than in the ordinary case of the creation of saviour siblings.

It seems, therefore, that a compelling case could be made for this course of action: the creation of a ‘saviour donated sibling’. Nevertheless, it is hard to avoid the suspicion that, at some level, the distinction between this scenario and the creation of saviour embryos is casuistry. If the therapeutic use of tissue derived from embryos is sometimes ethical, it is difficult to see how whether it is ethical or not could depend on whether or not another child had been born as the result of the process in which the embryo was created.

Therapeutic cloning 

Bioethicists have already extensively discussed an arguably much more powerful technology, which is related to the one under investigation here. ‘Therapeutic’ cloning would involve using somatic cell nuclear transfer to create a blastocyst from which to derive embryonic stem cells and then specific cell types that were genetically identical to the cells of the person being treated for transplant or other therapies. As noted earlier, recent advances in the creation of induced pluripotent stem cells suggest that it may eventually be possible to derive patient-specific stem cells for therapeutic purposes from somatic cells without the need to involve a human ovum or to create a blastocyst, in which case the justification for therapeutic cloning would lapse. However, what is important for the purposes of the argument here is the fact that numerous discussions of the possibility of therapeutic cloning have argued that it would be justified if it offered real therapeutic benefits.

Therapeutic cloning might be argued to be more ethical than the creation of saviour embryos on the grounds that it does not involve conceiving new embryos by fusing spermatozoa and eggs but only involves making copies of a genetic blueprint that already exists and could be further copied as required. However, this intuition relies on a misconception that genetically identical embryos would grow up to be the ‘same’ person. The creation (or destruction) of cloned embryos is no different to the creation (or destruction) of ordinary embryos – at least as far as the moral status of the embryos is concerned (Savulescu, 2000). Identical twins do not have a lesser moral status by virtue of being clones; neither should embryos which are genetically identical to each other or to an existing person.

Thus, like the creation of saviour embryos, therapeutic cloning would involve the creation of embryos with the intention of using them for therapeutic purposes. If therapeutic cloning would be ethical, so too would be the creation of saviour embryos. Of course, those who object to the creation of embryos for purposes that would involve destroying them are unlikely to hold that therapeutic cloning would be ethical, so, as was the case with the observations above about abortion, the argumentative traction of this analogy with those entirely opposed to the instrumental use of embryos may be limited. Nevertheless, the analogy with therapeutic cloning does establish that the creation of saviour embryos would be no more radical a step than another proposed technology with numerous defenders and substantial popular support (Devolder and Savulescu, 2005, Savulescu, 2000, Tooley, 2006).