Unethical Perinatal HIV Transmission Trials Establish Bad Precedent

By Udo Schuklenk

[This is a slightly revised version of a paper published in Bioethics 1998; 12:312-319. The references are omitted in the version on this site! U.Sch.]]Professor Resnik defends a number of clinical trials which took place in developing countries in order to test zidovudine as a means of reducing the perinatal HIV transmission rate. There are several ethical problems with these trials 1. Is it ethical to withhold a proven intervention such as zidovudine from research subjects in developing countries during placebo controlled trials, because, so goes the ethical justification, without the trials none of the research subjects would have been able to access this drug at all?2. Have the actual trials adhered to basic ethical research standards in regard to voluntary informed consent of the research subjects as a conditio sine qua non?3.1 Given that the purported objective of the trials was to develop a successful regimen of perinatal HIV transmission prevention which is affordable to women in developing countries, has this research actually achieved this objective? 3.2 Was the strategy of researching zidovudine as a first-choice drug in perinatal HIV transmission in countries such as Ivory Coast and Uganda looking at the most cost-effective solution from the outset (i.e. even if we would accept the above mentioned scientific rationale at face value)?

In two earlier responses to these trials I have maintained that the placebo controls were ethically unacceptable, but that efficacy tests of lower than ACTG 076 levels of zidovudine, with the objective of developing a cheaper drug regimen, are ethically defensible. I was wrong. I have changed my mind largely for the reasons provided in this commentary.Despite Professor Resnik’s explicit reliance on principlism, it seems to me that his argument is essentially of a consequantialist nature. It takes into account the interests of those women affected during the course of the trials, and those women in developing countries who are likely to benefit should the trials succeed in developing a cheaper drug regimen. He argues that trials under consideration were ethical, because the women in the trial would otherwise have had no access to the potentially sucessful HIV intervention zidovudine anyway, and because of the likely benefit to a very large number of HIV positive pregnant women. I agree with Professor Resnik’s philosophical basis for justifying the trials, yet I have different perceptions of the real world of these trials than he has. Given that consequentialist reasoning depends to a large extent on empirical information, it is of considerable importance to get these vital facts straight. In this commentary I will outline my more critical assessment of these trials on the basis of brief analyses of the above mentioned four ethical problems.

1] Placebo controlsI am still unconvinced that placebo controls were necessary. Not being a scientist, I have to rely largely on the judgement of a number of eminent scientists who conduct research clinical AIDS trials, and who have assured me that placebo controls were unnecessary (some of the reasons for this can be found in the article by Lurie and Wolfe). In fact, a strong indication supportive of this judgement can, ironically, be derived from the actual outcome of the trials, which confirmed the beneficial effects of zidovudine as a means of preventing perinatal HIV transmission. Taking into account information I received since publication of my two defences of these trials, I am now much more doubtful about whether they should have ever taken place in developing countries.

2] First person voluntary informed consentProfessor Resnik refers to sources close to UNAIDS as evidence for that the relevant ethics committees have been consulted, that eminent Western bioethicists have attended UNAIDS sponsored meetings, and so on and so forth. Given that UNAIDS had a vested interest in the realisation of these trials, one has reason to be sceptical about the types of experts it consults. The organisation has been supportive of these trials because it considered them a means to develop an affordable means of reducing the perinatal HIV transmisison rates in developing countries. The question ought to be asked whether UNAIDS’ ethics experts were likely to be willing to sacrifice internationally recognised research ethics standards? Did they support arguments put forward by Ijsselmuiden and Faden or Ekunwe, which require very high standards of first person informed consent, or were they prepared to go for the easier option propagated by Christakis? Christakis argues basically that consent of a community representative might suffice if first person informed consent isn’t feasible. Ijselmuiden, Faden and Ekunwe won’t have any of this, largely because they are doubtful of the legitimacy of many such ‘representatives’ in particular in developing countries with a poor human rights record. Nigeria based Ekunwe is adamant that "if the proposed tests are different from routine tests, and carry risks of complications and/or pain, then informed consent must be obtained either verbally or in writing." Evidence that many research subjects in developing countries have not understood what they consented to is mounting not only in currently ongoing UNAIDS supported preventive AIDS vaccine trials, but also for the trials under consideration by Resnik. Phanuphak reported in the New England Journal of Medicine that HIV positive pregnant women who participated in these trials "who sought antenatal care in early pregnancy were not offered complete information or told about the full benefits of the ACTG 076 regimen but, rather, were left untreated until the 36th week of gestation, when they were randomly assigned to the zidovudine or placebo group." If this happened in a developing country such as Thailand, which is a functioning democracy, one must inevitably become concerned about the treatment of research subjects in some of the other participating countries, which have a quite appalling human rights record. An editorial in The Lancet suggests that it is "insulting to imply that approval from an African ethics committee is valueless." Certainly, as a general statement this is correct, yet even Asian and African scholars admit that one of the reasons for conducting trials in developing countries is that often it is easier to gain ethics committee (if there is one at all) approval. Questionable is also the voluntariness of the HIV positive women’s consent. Isn’t it a coercive offer to force terminally ill pregnant women to choose between joining a placebo controlled trial which gives a shot at an established HIV intervention, and no treatment at all? In Western countries such patients have voted with their feet and gone to great length to join such trials only in order to break subsequently the trial protocol. They see trial participation as their only chance at a possible HIV treatment (vs. the placebo which is not even offering this chance). Marc Lallemant, Max Essex and colleagues commented on the zidovudine trials that "clinical research worldwide would be jeopardised if potential participants feared that, within a trial, they risked being denied the most effective treatments available." I think such trials constitute unethical clinical research, irrespective of their scientific rationale. There is no evidence that the trials Professor Resnik defended offered clinical equipoise between the placebo and the zidovudine arms. Similar trials take place currently sponsored by the French INSERM/Agence Nationale de SIDA in Ivory Coast. The objective is to determine whether a drug known in Western countries to prevent pneumonia in people with AIDS, works better than a placebo. The outcome of these trials is as predictable as the outcome of the zidovudine trials, yet AIDS patients in Ivory Coast will pay with their lives for knowledge that is unlikely to surprise anyone. One might argue that patients in these trials are better off than those who were not offered a chance to participate in them, because they were at least given a chance at an established means of preventing HIV transmission or, as in the Ivory Coast trials, pneumonia. At a first glance this argument seems convincing, yet it ignores that the international patent regime, designed to safeguard the economic interests of shareholders of Western pharmaceutical companies, has led to this situation in the first place. In other words, a coalition of Western governments and Western pharmaceutical companies has created a situation that effectively prevents the vast majority of people with AIDS (or other life-threatening illnesses) in developing countries from accessing affordable medication. Subsequently researchers sponsored by these very same governments conduct clinical trials in developing countries which ‘offer’ 50% of the participating trial subjects no treatment at all. To qualify this as ethical, because the other 50% receive an already established medication would open the door to conduct many more trials in developing countries which would be considered unacceptable in developed countries.

3.1 and 3.2] Viability of the economic justification for the trialsProfessor Resnik suggests a further ethical justification of the perinatal HIV transmission trials which is of an economic nature. The argument is essentially that it is legitimate to provide substandard treatments to people in countries where the standard of care is lower or at the level of what would be considered the substandard treatment arm in a Western society. This is, of course, only as long as there is some subsequent benefit for the people in societies which took risks in these trials. This is precisely where the crux of this argument lies. Utilitarians could accept this reasoning because they’d find it justifiable to sacrifice the interests of some few trial participants (who have given informed, and voluntary consent) for the sake of a much larger number of people who might gain as a consequence of this sacrifice. Three questions need to be addressed in this context:

a] Have the trials under consideration actually produced such a result, i.e. an efficient perinatal HIV transmission affordable to HIV infected women in societies such as Ivory Coast and Uganda? b] Would it have been possible to achieve similar results by means of researching alternative means of preventing perinatal HIV transmission? c] Given the substantial costs involved in preventing perinatal HIV transmission, is it ethical to spend these amounts of money on a drug for perinatal HIV transmission, given that in many other areas the same funds could be used more effectively to prevent HIV transmission?a] The answer to this question is ambiguous at best. The producer of zidovudine, UK pharma giant Glaxo-Wellcome has announced that it will offer this drug to women in developing countries at a price of between US$50-150. A drug with this price tag attached to it will be affordable to a limited extent in countries like Thailand, but it will almost certainly be out of reach for the majority of women in African countries, India, China, Burma etc. Even in some of the economically more advanced societies of Southeast Asia it is likely that zidovudine will be available only to the wealthiest HIV positive pregnant women, rather than to all HIV positive pregnant women who wish to make use of it. In Mexico, in 1997 the costs of a monthly regimen of (the known to be effective) combination therapy of AZT, 3TC and Indinavir is US$783 per month. This is more than eight times the minimum wage in that country. These drugs have been tested in clinical trials (approved by research ethics committees etc.) in Latin America, yet in reality they are unavailable to most people with AIDS who live there. Professor Resnik’s ethical-economic justification for conducting these trials is, at best, a weak one. In violation of the CIOMS research ethics guidelines no guarantees have been given that people in either of the countries where the trials took place will be offered affordable access to AZT. The drug manufacturer’s ‘discount’ translates in the real world into a marketing campaign rather than into affordable access. Waldholz reported in the Wall Street Journal that "even if AZT becomes affordable, many health officials doubt the drug will reach poor women any time soon. Few places, especially in Sub-Sahara Africa where the epidemic is raging, can afford the medical facilities or staff to identify infected women and treat them daily for three to four weeks before they give birth." He expects another US$ 60 million a year to be added to Glaxo-Wellcome’s revenue as a result of its ‘discount’.b] Michelle Murrain, a feminist author and neurobiologist, has vigorously criticised ACTG 076. She points out that zidovudine isn’t really as efficient as people seem to think. What this trial really told us is that infants of HIV infected women have a ³ 75% chance of being born uninfected in any case. In some countries the transmission rates go below the 10% mark, while in others they are as high as 40%. Clearly there is a strong correlation between the general health of HIV infected pregnant women and the transmission rate. No research has been undertaken as yet to determine which factors exactly contribute to decreases/increases in HIV transmission rates. Ridge and Arachne have pointed out that the interest of Western sponsors in the actual health needs in developing countries is very limited, and support for research on traditional methods which might help to reduce HIV transmission rates is non-existent, largely because there are no economic gains to be made. The current strategy is to hook people in developing countries onto extremely expensive Western drugs rather than to research solutions adequate to the conditions of people in those countries. The answer to the question I posed is therefore that it is unclear whether it would have been possible to achieve better results, because the necessary research has never been undertaken.As to c]. A recent study of 12,537 adults in Tanzania, for instance, indicates that a much more cost-effective strategy to reduce the incidence of HIV is to avert HIV infection by means of improving the management of sexually transmitted diseases by primary-health-care workers. To put it quite simply, if we accept at all that economics is a factor in determining whether the trials in question were ethical, we need to determine whether the means they suggest to prevent HIV transmission are cost-effective not only when compared to the price of the intervention in Western countries, but also within the context of the overall funds available, and within the context of other means designed to prevent HIV transmissions. Here zidovudine constitutes not a particularly cost-effective means, if we compare it to alternative strategies. This is important in societies which have to ration the allocation of resources within their respective health care budgets. In Thailand, for instance, the government had to decide whether to spend the money available for AIDS primarily for education (prevention) or for actual healthcare. In 1997 75% of its AIDS-related funds went into healthcare while 25% were allocated to HIV prevention. This meant that the funds available for healthcare for people with AIDS were grossly insufficient from the outset, and decisions had to be made about the level of healthcare that should be provided. Zidovudine has been introduced through the backdoor of these clinical trials. People in Thailand are now aware of its existence and will demand access to it. The consequence is that expenditure in other areas will have to be scaled back, even though the prevention of HIV transmission might have been more cost-effective there. Much of this has to do with observations made by Solomon R. Benatar, "Market driven healthcare in a powerful nation has profound symbolic and material effects on health services worldwide. Health delivery is being transformed ... toward the sale of a commodity in an adversarial marketplace. ... Legitimation of this transition transmits a powerful message that gravely undermines the will and ability of (developing) nations to develop and sustain equitable access to affordable health care." Looking at the trials as Professor Resnik did, that is, without taking seriously into consideration the socio-political context in which they took place, leads to inadequate answers.