Availability after a trial

We have seen that in the current guidelines there is a requirement that the drug or vaccine being tested should be made "reasonably available" after the trial. Examination of an illustrative case will show that there are reasons why this should not be a general requirement for research in developing countries.

Trial of chloramphenicol for meningitis in Nigeria

In the beginning 1970s treatment of meningococcal meningitis was resistant to treatment with sulphonamides, a cheap and effective treatment. Another treatment, intramuscular injection of penicillin, was also found to be inapplicable, because of its expense and difficulty of administration. The trial attempted to find out whether chloramphenicol, given alone, a much cheaper treatment than penicillin, would be effective against meningitis. The trial showed that it was indeed as effective as penicillin (Whittle et al. 1973).

This relatively simple case illustrates, I think, the problems with demanding a written contract of availability. Existing treatments of meningitis were unsatisfactory because of costs and logistics. The researchers therefore had a clear justification for seeking a cheaper and more practical alternative, and a reasonable expectation that the treatment would be a viable alternative if found to be as effective as penicillin. For the sake of argument, I do not here question that medical judgement. To demand, in addition, that the researchers, as a condition of initiating the trial, secure an agreement with relevant authorities that the treatment be made available to the communities in which it was tested, would go far beyond the influence one can reasonably expected researchers to have concerning changes in a country's health policy. Instead, it will probably have the effect of delaying or preventing research that might benefit the population.

One should also note that there are formidable obstacles to achieving such agreements before trials start. Not only are there issues of intellectual property rights, but there are also all kinds of logistical reasons that need to be solved before a drug or vaccine is made available in the country in which it is tested. New drugs and vaccines are going to be expensive. Although there are several proposals for differential pricing schemes, they have not been successful so far. But even if drugs were free this would not ensure availability. There needs to be an appropriate health care delivery system, often absent in developing countries. The story of how hepatitis B vaccination was introduced into the EPI program illustrates the difficulty of making effective drugs or vaccines available to populations who need them, difficulties which go far beyond the reasonable responsibilities of the sponsors of the trials (Muraskin 1995).

Considerations such as these show that the CIOMS requirement of an agreement to ensure the reasonable availability of the product being tested, is probably not a viable option. An alternative guideline that still captures our intuitive sense of justice is therefore urgently needed. But let us before we do that, examine another option, namely that we should not be concerned about availability at all when initiating trials in other countries.

No obligation to ensure availability: The hepatitis A vaccine trial in Thailand

This trial started in Northern Thailand in 1991 and involved approximately 40.000 children, from 1-16 years (Innis et al. 1994). The hepatitis A vaccine was Havrix produced by SmithKline Beecham. The study concluded that "inactivated hepatitis A vaccine is safe and that it protects against hepatitis A for at least 1 year". The justification for the vaccine trial in Thailand was the following

"In studies conducted with the Ministry of Public Health of Thailand between 1987 and 1989, it was concluded that access to a safe, affordable, effective hepatitis A vaccine was a future public health requirement. Accordingly, an evaluation of the safety and then of the protective efficacy of hepatitis A vaccine was planned; primary schoolchildren were to be the study participants. Kamphaeng Phet Province in northern Thailand was selected as the filed site because we have identified, by active surveillance between 1989 and 1991, this province as the site of an attack rate of hepatitis A of 199 cases per 100.000 population. Moreover, our experience in using the province as the site of a large field efficacy evaluation of Japanese encephalitis vaccine led us to believe that the people of Kamphaeng Phet would welcome hepatitis vaccine study"

The hepatitis A vaccine has been licensed by SmithKline Beecham, and is now mainly used for travellers from industrialized countries to developing countries. Although there is some discussion about whether one should introduce routine vaccination programs among high risk groups, it is seems that the general consensus is this is not cost-effective at this time. Thus it should not have been too difficult to predict that the vaccine would not become generally available in northern Thailand after this trial had been completed. The main reason for doing the trial was to get the vaccine licensed in industrialized countries. The population in Thailand was a convenient one, with a high incidence of hepatitis A. This trial, then, is an example of a trial done in a developing country, where it was known that the product would not be available to the population after testing, and the results would primarily benefit people in the developed world.

Some might argue that there is nothing wrong in carrying out such a trial under these conditions. After all, those who take part in the trial are not harmed; those who receive the active vaccine obtain some benefit, the country obtains some benefit by taking part in international research, and the trial produces knowledge which will benefit a large number of other people, and will advance knowledge in general. Why, then, would it be wrong to carry out a trial in a country, with the intention that the results will primarily, or exclusively, benefit people in other countries? If we feel that somehow the availability question needs to be adressed, we must give an answer to this challenge.

The challenge is indeed a difficult one. One might try to say that the participants in the trial are exposed to risk, or that they would not consent to a trial, the results of which would benefit other people. But that does not show that a trial where there is an appropriate risk benefit ratio, as there should be, and where the participants are properly informed about the possible use of the results of the trial, as they should be, is morraly wrong. One might argue that participation in an international research project for the researcher is not enough to justify the trial, but that again only shows that if an appropriate compensation had been provided, the trial would have been justified. None of these arguments shows that there is anything in principle wrong with doing a trial in one country for the benefit of people in another country.

The trial caused quite a bit of a controversy when it became a matter of public record in Thailand, showing that there is something about this situation we feel is morally problematic. I would suggest two explanations for that. It could be that people felt that this particular trial was not justified in light of all relevant knowledge about it, including the balance of risks vs. benefits to participants and to the country in general. Or it could be that the public controversy was caused by the fact that there had not been full disclosure and discussion about the appropriateness of deciding to this particular trial in Thailand.

What this shows is that, in the same way as it is problematic to justify a general requirement that one should have a contract concerning reasonable availability, it is also difficult to justify a general prohibition against trials carried out in one country for the primary benefit of people in other countries. A middle course is what we should be looking for.

Share in the profits of the results of research

One might be tempted to propose that it would be reasonable for the country or community to receive some share in the profits should the drug or vaccine prove to be effective. After all, the community has contributed something to the development of the drug or vaccine, and it seems only fair that it should also take part in the profits.

There are several problems with this approach, which does not make it an attractive option. The first concerns how one should quantify the exact contribution of the clinical trial in drug development. It is, of course, absolutely essential to do a phase III trial before regulatory approval; in that sense the trial does make a vital contribution, giving the investigators or the country some bargaining power. It also remains true, however, that a particular trial site is not going to be essential for the development of the drug, any appropriate trial site will do, thus lowering the bargaining power of any individual site. One could, of course, counter that there should be some international agreement on an appropriate percentage of the profits as compensation for the contribution of a clinical trial site. While not impossible, at present this would require a radical rethinking about our approach to intellectual property rights of pharmaceutical products. It would be more in line with current thinking for the sponsor to pay for services rendered, i.e. to pay up front whatever the clinical trial is worth for the company. This is, of course, what is done right now, although it probably is possible to demand a higher payment than is done today; i.e. not only payment for direct expenses and some infrastructure, but some additional reasonable payment over and beyond these costs.

This somewhat negative assessment of requiring a percentage of profits needs to be modified somewhat, in light of the fact that the International AIDS Vaccine Initiative (IAVI) has been able to secure such an agreement for the vaccine candidates it is sponsoring, although the agreement seems to be limited to patents filed by academic centers.

"All participants in IAVI''s vaccine development partnerships agree to an intellectual property agreement that allows both for free-market recouping of investments and for faster and wider distribution of vaccines. Under the IAVI intellectual property agreements, product developers can charge market prices in industrialized countries. When patents are filed by an academic center or its partners, however, IAVI is free to choose from one of several measures to ensure global accessibility: a non-exclusive, royalty-free license for production in the developing world; an exclusive license with certain restrictions; 25% of all net royalties or sales from intellectual property rights; or an agreement from the manufacturer to sell vaccines at a small percentage above cost in developing nations."

Another approach which is being explored in a number of different settings is to establish private-public partnerships, such as the drugs for malaria venture, UNAIDS AIDS drugs scheme in countries such as Uganda, or attempts to establish funds that would cover the costs of an HIV vaccine for poor countries when it becomes available. There is no doubt that such schemes are important and should be encouraged. However, it remains to be seen whether they really are going to be effective at solving the availability problem. The difficulties encountered also remind us what a challenge this problem is, and of the need to go beyond the rhetoric of "reasonable availability" and to propose achievable mechanisms to ensure availability of new drugs.

Some illustrative examples of past research in developing countries

There are thus good reasons to be sceptical of an approach that demands a contractual guarantee of access to investigational products after successful testing. Nevertheless, the phrase "reasonable availability" does point to something we desire out of a trial. I shall now show that research in developing countries is often justified in terms of a desire to make useful products available to the population in which they are being tested. More often than not, however, these justifications do not materialise after testing is completed, and this could have been known at the start of the research. I shall therefore suggest that, if one thinks availability is a condition for testing, much more explicit and detailed justifications need to be provided. The problem now is that researchers have an overly optimistic attitude towards the benefits which the research will bring. It is often simply assumed that if a treatment is found to be effective, it will be made available to the populations who need it. The examples clearly demonstrate that this is generally not the case, but also that there are different types of scenarios in terms of expected availability. All of this leads me to conclude that the way to go is to demand a much more explicit and detailed justification for the trial before initiation in terms of benefits to the community, rather than a general rule prohibiting or requiring certain types of trials.

Short course treatment of tuberculosis

In 1972 a controlled trial in East Africa on short course treatment (6 month) of pulmonary tuberculosis was carried out(East African/British Medical Research Council 1972). It had been found hard to ensure compliance with the standard 18 month treatment regimen, both in developed and developing countries. The authors recognize, that "it was fully appreciated that the regimens which were selected for study were not readily applicable under routine services conditions because the involved daily streptomyocin for 6 months." The study proved that short course treatment were indeed effective. However, the authors also conclude that "it is now important to try and find short-course regimens that readily lend themselves to practical application by routine treatment services. This is necessary not only in technically advanced countries but, even more important, in the developing countries where the whole organisation of programmes of long-term chemotherapy has repeatedly failed and has usually proved to be beyond the capacity of the routine treatment services as currently financed, organised, equipped and staffed. Further studies are being directed to these ends". In other words, it was recognized that this particular short course treatment would not be practically implementable in the country in which it was carried out, although the trial was necessary to establish that short course treatment was effective, a necessary first step towards the development of applicable treatments. The same is the case for the next case:

Malaria prophylaxis among pregnant women at the Thai-Burmese border and rural Malawi

A number of studies have been carried out to investigate the effects of malaria prophylaxis during pregnancy. One of the interests has been to see whether the drug mefloquine has any adverse effects on the child. One such study was carried out at the Thai-Burmese border (Nosten et al. 1994). The other one investigated such things as interuterine growth and birth weights in Malawi (Steketee et al. 1996). Mefloquine is an expensive drug that is not generally available for malaria treatment in the developing countries where malaria is a major problem. It certainly would be too expensive for routine malaria prophylaxis during pregnancy. According to Molyneux, the results from the Malawi study were used to apply for approval in the US for the use of mefloquine during pregnancy, although this application was denied (Barry and Molyneux 1992). Molyneux also argues that trials such as these are justified; although the treatment will not be immediately available after the trial is completed, it is a necessary first step towards the eventual availability of treatment.

Zidovudine treatment for pregnant women in a sub-saharan country

In a recent article on the proposed revision of the Helsinki declaration, a study involving pregnant women is described (Brennan 1999). The women would not receive AZT to prevent perinatal transmission, although this has been found to be highly effective:

"Our debate about the proposed research involving pregnant women in a sub-Saharan African country was quite animated. As the discussion progressed, it became clear that the standard of care was changing in this particular country, where programs to provide zidovudine is changing in this particular country, where programs to provide zidovudine to pregnant women will be implemented in urban centers over the next six months. The researchers believed there was a small window of opportunity to conduct their research, because rural programs for the distribution of zidovudine will be introduced within two years. A number of committee members found this argument persuasive, given the value of the proposed research.

Others countered that the increasing availability of zidovudine should be a reason for not proceeding with the research. They also pointed out that the research team itself could probably provide zidovudine to pregnant women in the areas where the study would be conducted. Not to do so, they argued, would be to allow the transmission of HIV infection to children, even though it could be prevented - a violation of research ethics.

The reasoning by those who defended this trial is illustrative. First it is pointed out that there is a unique opportunity to do the study in this population (the details of the study are not disclosed in this article), as it is a population which does not currently receive AZT. Second, one must take advantage of this opportunity now, as AZT is in the process of being introduced in the country. Postponing the study will make it impossible to carry out the study. The problem with this reasoning is that it gives the researchers a potential and real conflict of interest. They will be strongly motivated to delay the introduction of AZT treatment in their country until the study is finished. Given the few number of experts in sub-Saharan countries, the researchers are also probably the advisors to the Government on AIDS treatment policies.

A breast cancer study in Vietnam

Love and Fost (1997) discuss a controversial study in Vietnam where women with breast cancer are randomized into one group receiving mastectomy and the other group receiving mastectomy together with immediate surgical oophorectomy and tamoxifen for the next 5 years(Love and Fost 1997). Adjuvant treatment for breast cancer patients is generally felt to be superior to mastectomy alone, although there is still scientific doubt as to the benefits of such adjuvant therapy. There is no doubt that it would be useful to carry out a properly controlled trial, although as the authors point out, such trials are "not politically feasible at present" in the US.

The principal investigator, Richard R. Love, has extensive experience from Vietnam and knows that adjuvant therapy is not generally available in that country. He therefore proposes that such a trial should be carried out there. The trial would have the added benefit of serving "as a means for technology transfer, education and immediate improvement of treatment." The women in the trial would not be denied treatment they would ordinarily receive, and in fact would have the chance of receiving a treatment that might be superior.

The ethics committee in the US initially rejected the proposal based on the following:

"The proposed study includes 2 treatment approaches, neither of which is standard treatment in the investigator's country, the United States. One arm - oophorectomy and tamoxifen - at least has the virtue of being a plausible approach for adjuvant treatment, worthy of investigation under acceptable conditions. The other arm - no adjuvant therapy after surgery - should be considered inadequate anywhere. It is exploitative of the women in Vietnam not to provide some systemic adjuvant therapy."

The committee nevertheless eventually approved the protocol with some modification, after the investigators had argued their case. The women in the mastectomy only group would be offered oophorectomy and tamoxifen if cancer recurred, and offered follow up visits every 3 months. This was based on the following arguments:

"A woman who was randomized to the observation group - no adjuvant treatment - would be no worse off a result of being in this study. The study would expose such a woman to no risk beyond that to which she would have been exposed were she not in the study. On the contrary being in the observation group offered the possibility of substantial benefit, namely more careful follow-up and planned treatment, standard for recurrence of cancer, if and when that should occur."

There is apparently some reason to believe that the adjuvant therapy in the proposed study in Vietnam are more risky than those currently regarded to be standard treatment in the US. Nevertheless, the investigators felt the study to be justified on the grounds that:

" For a population at very high risk for recurrent disease because of the the limited effective care available, higher risks might be appropriate. .... Given the poor prognosis of breast cancer in a country where no adjuvant treatment was available, it would not be irrational for a woman to assume a risk of adjuvant therapy that might be higher than that from a standard treatment in another country."

The investigators also argued that all studies on breast cancer treatments had been done in European and American women, and there is no data on the effect of adjuvant therapy on Asian or Vietnamese women. It is argued that the goal of the trial is to investigate treatments of relevance to Vietnam:

A crucial point in this aspect of the discussions with the IRB was clarifying that the research project had to be understood as part of a broader context, namely, to define an effective treatment for the most common presentation of breast cancer in Vietnam, which could be widely applied in that country. It was also important to convey the larger health priorities in Vietnam, particularly the control of communicable diseases. In this context, the investigator's proposal was considered consistent with international guidelines for research in underdeveloped countries in that it was "responsive to the health needs and priorities of the community in which it is to be carried out."

I must confess I find it really hard to understand this reasoning, and I suspect that the real reason for wanting to do the trial is not to benefit the population in Vietnam. There is clear scientific justification for wanting to do a clinical trial on hormonal adjuvant therapy; even though the study is done in Vietnamese women, there is no reason to expect that the results will not be seen as relevant for hormonal adjuvant therapy anywhere. It is, however, not possible to do the trial in an industrialized country. Any arguments that this trial is of particular benefit to Vietnamese women should therefore be viewed with suspicion. No data is given to support the claim that this is a treatment that "could be widely applied in that country". If the argument holds that it is necessary to do a study on Asian women to prove that adjuvant therapy is beneficial in this population, then one should choose one of the other types of adjuvant therapy, which, it is pointed out, probably carries a lower risk than the one in this study.

The short course AZT treatment of pregnant women

The final example is the very controversial trials on short course treatment to prevent perinatal HIV transmission. The main justification for these trials was that they were necessary to find a treatment that could be applicable in developing countries. However, it could have been known before the trials started that even a 50 dollar treatment would be too expensive for most, if not all, countries in which these trials were carried out.

Lessons from the cases

These cases illustrate two important features of research in developing countries that I think one needs to pay attention to when suggesting appropriate principles of justice.

First, a guiding principle in the selection of an appropriate research setting is the need to select an appropriate population for the sake of the scientific question being asked. The needs of the community in which the product is being tested usually takes second place. If the product being tested needs to be approved by regulatory authorities in a developed country, it is that concern that will be the determining factor in the choice of research population and research design. This is the case in the hepatitis A trial and the mefloquine trial during pregnancy. Even if regulatory approval is not the issue, the desire to obtain generalizable knowledge, as in the case of the breast cancer trial in Vietnam, dominates.

Second, there are always attempts to provide a justification of the trial in terms of benefits for the population in which it is tested. Even the hepatitis A vaccine trial, where there is no doubt that a vaccination program in the population is not a viable option, such justification is provided. The breast cancer trial in Vietnam also provides such a justification, although there are obvious gaps in the reasoning provided. In cases such as the AZT perinatal transmission studies, the justification in terms of availability is obviously inadequate.

One should note that the cases do not represent examples of risky research for the sake of future benefits to others. In most cases, the participants in the projects receive some benefits by taking part in the project, and one would probably even say that on balance the benefits to the subjects outweigh the risks to the same subjects. In terms of the ordinary evaluation of risks and benefits before the start of research, there are few problems with these projects. What is lacking is a more than rudimentary justification of the research in terms of benefits to the populations during and after the study. If one thinks availability is important, a much more explicit and realistic justification before the research starts is thefore needed. But, as I have argued above, I do not think availability needs to be a general precondition for such research. There should, however, be a realistic description of the benefits to be expected from the research, and not merely empty wishes.