Intended for healthcare professionals

Education And Debate

A tale of two novel transplants not done: the ethics of limb allografts

BMJ 2002; 324 doi: (Published 20 April 2002) Cite this as: BMJ 2002;324:971
  1. David Benatar, associate professor (dbenatar{at},
  2. Don A Hudson, associate professorb
  1. a Department of Philosophy, University of Cape Town, Private Bag, Rondebosch 7700, South Africa
  2. b Department of Plastic and Reconstructive Surgery, University of Cape Town, Observatory 7925, South Africa
  1. Correspondence to: D Benatar
  • Accepted 13 September 2001

When it becomes technically possible to perform a “ground breaking” surgical procedure, an important moral problem arises. In view of the strong incentive for surgeons to be the first to perform a novel operation, their judgments about whether such an intervention is justified may well be clouded. We describe two cases in which surgeons decided not to perform a novel transplant. In the second case transplantation continued to be considered until the ethical analysis performed in writing this paper was completed.

Summary points

At the Red Cross Children's Hospital in Cape Town it was decided, on ethical grounds, against novel limb allografts for two patients—one with no hands and one with no hands or feet

The costs to health and cumulative lifetime risks of immunosuppression for child recipients of a limb allograft are substantial and do not outweigh the functional benefits

There are good reasons for not using “informed consent” to bypass the difficulties of weighing the benefits of limb allografts against the costs

The ideal candidate for limb transplantation is somebody who could gain the most and for whom the associated costs and risks are lowest

A potential recipient should be very close to this ideal if transplantation is to be morally justified

Two amputees

In July 1994, a 3 year old boy was admitted to the intensive care unit at the Red Cross Children's Hospital in Cape Town with meningococcal septicaemia. He developed infarction of all his fingers and these were amputated at the level of the carpus bilaterally with skin grafting of the stumps.

In January 2000, a 4 year old girl was admitted to the same hospital, also with meningococcal septicaemia. She developed distal necrosis of all four limbs. Both hands were amputated just distal to the carpus, leaving three metacarpal stumps on the right hand and a vestige of one metacarpal on the left hand. The left leg was amputated above the ankle and the right leg through the talus. The residual defects were subsequently skin-grafted. Tests for HIV and hepatitis B virus gave negative results.

There is an experienced team of microsurgeons at the University of Cape Town and its teaching hospitals, of which the Red Cross Children's Hospital is one. Hand transplants were considered for the first patient and transplants for all four limbs were very seriously contemplated for the second patient. The first limb allograft took place in September 1998 in Lyons, France, and the second in Louisville, Kentucky, USA, in January 1999. A few others have been done since then, but had the Red Cross Children's Hospital surgeons decided to transplant in 1994, their case would have been the first, and the second child would have been the first to receive allografts of all four limbs. Notwithstanding the professional incentives to perform the procedures, the surgeons decided against them, preferring a more conservative policy. Here we defend that decision, focusing mainly on the second case.

Ethical assessment

The heart of the ethical dilemma in limb allografting is this: although there is much to be gained from such a procedure—the ability to perform activities of daily living—the benefits do not include saving or lengthening life. Thus, they do not obviously outweigh the risks and harms associated with immunosuppression. The benefits of limb transplants are primarily functional—acquiring the ability to hold and manipulate things, and thus to be able to dress, feed, and clean oneself, and the ability to walk. There are cosmetic benefits, but these are relatively minor in comparison with the functional benefits. Thus ethical assessment of limb transplants has to weigh functional benefits against risks and harms associated with immunosuppression.


All things being equal, the greater the disability, the greater the potential functional benefit from transplantation. A person with only one hand, for instance, is not nearly as impaired as somebody who has neither hands nor feet and is therefore incapable of independent living. In this regard the Cape Town patients stood to gain much more than the recipients of the Lyons or Louisville transplants, both of whom had lost only one hand.

In another regard the Cape Town patients stood to gain much greater functional improvement. The literature on limb replantation shows that restoration of function depends largely on nerve repair1 and postoperative use of the flexor tendon. The major factors affecting results are age2 3 4 5 6 7 8 9 and the level of amputation. 10 11

The youth of a patient does not strengthen the case for transplantation in every respect of functionality. Firstly, those who lose limbs as young children probably adapt more easily to their new circumstances than they would if they lost the limbs later in life. Thus the functional disadvantage is less for child amputees. Moreover, a younger recipient of a transplant is likely to live longer and thus stand a greater chance of eventually rejecting it. Indeed, in studies on large animals and primates, no allografts have yet survived for a long time.12 Giving a child a limb transplant could thus result in a relatively normally functioning limb for a limited time followed by the loss of that limb at a stage of life when adapting to its loss is more difficult. Another reason why the later loss would be psychologically and functionally more devastating is that the limb would have to have been cut back for the original allograft.

Risks and costs of immunosuppression

The risks and burdens of immunosuppression are well worth bearing when the alternative to transplantation is death. Matters are less clear when the alternative is a life of normal duration but without the benefit of one's limbs. Contemplating limb transplantation thus requires a much more careful assessment of the risks related to immunosuppression.

Malignancies are the most serious concern. Immunosuppressed patients are considerably more likely to develop certain malignancies, but the data on these rates have to be interpreted carefully. The average prevalence of malignancies in renal and cardiac transplant recipients is 6%.13 Skin cancer constitutes the largest proportion of these, but as it is nearly unknown in people with very dark skins, it would not be a risk for either of these two patients, both of whom were black Africans. Kaposi's sarcoma is also common in transplant recipients, but although sub-Saharan Africans are at increased risk, it has been found to be uncommon in children who have had transplants.14

Lymphomas are the second most common malignancy among recipients of transplants. However, as those who performed the transplant in Lyons noted in assessing the risks to their patient, the “risk of B-cell lymphoid tumour … probably stays at less than 1%.”15 This ignores the lifetime cumulative risk. The patient in Lyons was 48 and the patient in Louisville 37. Because composite tissue transplants require more rather than less aggressive immunosuppression and thus the risk of lymphoma is more likely to resemble the higher lymphoma risk after cardiac rather than after renal transplants, let us assume an annual risk comparable to that for heart transplant patients, which is 0.3%.16 Let us assume, secondly, that each patient has a life expectancy of 80. A 0.3% annual risk of lymphoma over 32 years for the Lyon patient and 43 years for the Louisville patient accumulates to 9.6% and 12.9% respectively over their remaining potential life spans. For the very young Cape Town patients, however, cumulative risks have to be computed for a much longer period—56 years for the second patient if we assume a life expectancy of 60 years. That is a cumulative risk of 16.2%. Some may wish to offset this against, firstly, the ordinary increase in risk of malignancy for immunocompetent people as they get older, and secondly, the fact that the later in life that a malignancy occurs, the fewer the years of life it cuts short. However, even considering this, the cumulative risk for immunosuppressed patients is still very substantial. Moreover, the risks of lymphomas after transplantation are considerably higher in children. 17 18 There should be serious concerns about inflicting such risks on a child who does not have a life threatening condition.

Some might be tempted to say that if a lymphoma does develop, it can be treated, even if that requires removing the limb and restoring immunocompetence. But this plan has shortcomings. Firstly, it is unlikely that a transplanted limb would readily be removed: given the effort that goes into transplanting it in the first place, doctors and patient would be inclined to attempt to tackle the disease without sacrificing the limb. Secondly, if the limb were to be removed, it would result, as we indicated earlier, in a more serious impairment and psychological loss than did the original amputation. Finally, there are no guarantees that after all this the patient's life would be saved.

Malignancy is not the only risk from immunosuppression after transplantation. Opportunistic infections are common. Although these can be managed, and the risks decrease after six months, many patients have enduring problems.12

Immunosuppressants themselves have adverse effects, such as nephrotoxicity, though this usually causes no morbidity and low mortality. Other adverse effects include hyperglycaemia, hypertension, hirsutism, and suppression of growth. Even when these effects are manageable and do not threaten life, many entail a toll of morbidity that is qualitatively different from that caused by absence of limbs. Instead of feeling healthy though handicapped by loss of limbs, transplanted patients experiencing such adverse effects are likely, although they have functioning limbs, to have a chronic feeling of compromised health.

Immunosuppressive treatments may improve vastly in the coming decades. However, if basing decisions on what might happen in the future were acceptable, then one could as easily argue in favour of not transplanting because it is possible that significant progress could be made in nerve regeneration research, thereby allowing better functional results for child patients when they grow up and receive transplants as adults.

Social circumstances

Another factor that affects the suitability of the second patient for transplantation is her poor social circumstances. She and her unemployed 27 year old mother live in a shack with electricity but no running water, no telephone, and only an outside toilet. In addition, transport expenses for travelling to regular medical appointments would be considerable. Such living conditions are obviously not ideal, but they are not a deciding factor against transplantation. Surgeons at Red Cross Children's Hospital have achieved some good outcomes with transplants of solid organs in children who live in very poor social circumstances.

Informed consent

Were there no disadvantages from delaying transplantation it would be preferable to await this child's own assessment of the disadvantages and benefits of transplantation when she becomes an adult. But there are disadvantages from delay. In such circumstances, (informed) decisions have to be made by custodians—usually the parents—of a child's interests.

Some people are inclined to bypass the difficulties posed by weighing the advantages of transplantation against the disadvantages, by arguing that whether the transplantation is acceptable depends on the choice made by the patient (or the proxy decision maker). Those who performed the Lyons transplant “believe that the recipient is the only person able to make an appropriate decision.”15 Although we agree that no wrong is done to a patient who freely gives truly informed consent to any procedure, two issues need to be considered before a full moral assessment can be offered.

Firstly, there is ample evidence that people's judgments can be affected by the way issues are presented—the so called “framing effect”—even when no such effect is intended. The mother of the second patient was informed of the benefits, the risks of death or rejection, and the need for lifelong drug treatment, and she was taken to the renal transplant clinic to see children who were taking immunosuppressive drugs. Nevertheless, it is impossible to avoid framing effects entirely. At that time the surgeons believed that, all things considered, transplantation would serve the child's best interests. The mother agreed and was willing to consent and sign the necessary forms. However, had the surgeons had a different overall assessment, as they now do, then the matter might have been either presented in a subtly different way to the mother or perceived differently by her.

Secondly, we need to distinguish between the acceptability of a person choosing some risky or costly course, and someone else facilitating that person's assumption of the risk or cost. It is because the second option may be morally problematic, even when the first is not, that doctors may be morally prohibited from offering or agreeing to provide a “therapeutic” option that they reasonably deem to be too dangerous. This is true when dealing with competent patients and even more so when proxy decisions are made for patients who are not competent. The Cape Town surgeons ultimately judged the risks of transplant to be too great.


The ideal candidate for limb transplantation is the person who could gain the most and for whom the associated harms and risks are lowest. But just how close to the ideal must a potential recipient be before transplant is acceptable? We have argued that the answer should be “at least very close.”


We are grateful to Wendy Spearman and Charles Swanepoel for discussing the immunosuppression issues with us. Wendy Spearman also provided some helpful references.


  • Competing interests None declared


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
View Abstract

Log in

Log in through your institution


* For online subscription