Ethics, antiretrovirals and prevention trials

By Emily Bass*

The recent decisions by several AIDS vaccine trial sponsors to ensure access to antiretrovirals (ARVs) for trial participants who become infected with HIV come after years of debate and discussion about the ethical implications of providing—or not providing—these powerful medications to trial volunteers.

Two of the central questions were: Is there an ethical obligation to provide ARVs to volunteers who become infected during the trial period through high-risk behavior? Is it unethical to provide these medications solely to a segment of the population, that is, trial volunteers, when other community members cannot obtain them? Perhaps surprisingly, the new policies do not provide definitive answers to these questions. Instead, they propose distinctions between a trial sponsor's ethical obligations to volunteers and their options for providing healthcare benefits.


Many trial sponsors say that trial design considerations, not ethics, were a primary motivation for the decision to include ARVs in trial-related benefits. HIV Vaccine Trials Network (HVTN) head Larry Corey says that the need to learn about how vaccines impact on HIV disease in infected volunteers was a key consideration. Ensuring ARV access increases the likelihood that volunteers will remain in the study, and compensates them for their participation in a lengthy follow up period. “This is the right thing to do both scientifically and operationally,” Corey says. Seth Berkley, President and CEO of IAVI, takes a similar view. “In large part this is a protocol issue for AIDS vaccine trials; it is not a statement about what is or is not ethical.”

These pragmatic rationales have the most bearing on large-scale efficacy trials which will ask HIV-infected volunteers to participate in long-term follow-up studies. However, sponsors say that they will also extend the benefit to participants in small-scale trials.

These decisions have not supplanted the ethical considerations raised over the past several years. HVTN, IAVI and the South African AIDS Vaccine Initiative (SAAVI) all developed their policies in consultation with ethicists after review of the existing ethical guidance documents on research in human subjects. In South Africa this process led reviewers to the conclusion that, based on strict interpretation of ethical principles, vaccine trial sponsors are not obliged to provide ARVs to research participants who are not infected with HIV at the outset of the trial and who receive high quality prevention counseling throughout. “It was hard for our committee to see why there was an ethical compulsion to provide ARVs,” says researcher Cathy Slack (University of Natal), a member of the South African HIV AIDS Vaccine Ethics Group. “But it was possible to make a strong argument that providing ARVs was morally praiseworthy—an act of positive beneficence.” 

This notion of ARV treatment as praiseworthy, rather than mandatory, is an important distinction to AIDS vaccine trial sponsors, as well as to sponsors of other prevention trials. While some international guidance documents do address the issue of AIDS vaccine trial sponsors’ responsibility toward volunteers who become HIV infected during the trial, none provide a clear directive on the issue. The 2002 CIOMS (Council for International Organizations of Medical Sciences) guidelines state that provision of treatment for the disease that a vaccine is designed to prevent is “morally praiseworthy” but not mandatory, and recommends that decisions be made in consultation with the host country and communities where research will occur. The WHO UNAIDS ethical guidelines for preventive AIDS vaccine trials state that at a minimum sponsors should provide the “highest level of care attainable in the host country.”

AIDS vaccine trial sponsors are aware that their actions will impact on other fields of prevention and perhaps lead to new expectations of all areas of research. This is particularly true for microbicide trials in which participants who become HIV infected are not followed for long periods of time since there is no presumption that microbicides will alter the natural history of infection. Microbicide researchers and advocates participated in many of the discussions leading up to the vaccine trial sponsors’ new policies. While there is growing consensus within the microbicide field that trial sites and sponsors will advocate for scale up of ARV access programs that will benefit volunteers, they have not yet committed to fund these programs themselves.

“Treatment should occur for all AIDS vaccine trials, but what do we do about other prevention trials, will we paralyze them? That’s a great fear in my mind,” says Berkley. “You have to be very careful not to condemn the future,” says Kevin DeCock, head of the US Centers for Disease Control and Prevention (CDC) program in Kenya.  “If you’re very prescriptive about things, you can end up backing yourself into a corner. You can’t measure the effects of research that doesn’t get done.”

As trial sponsors begin to implement their new policies, ethicists continue to explore the principles and implications of these decisions and how they might affect other scenarios surrounding AIDS vaccine trials. A recent Clinical Trials Subcommittee meeting at IAVI to consider the implications of such healthcare provisions revealed disparate views among some of the experts in attendance, prompting IAVI Report editor Simon Noble to conduct an on-line debate between bioethicists Ruth Macklin and Charles Weijer. Their strong, sometimes controversial positions are illustrative of this dynamic field.

Ruth Macklin, PhD, is Professor of Bioethics in the Department of Epidemiology and Population Health at the Albert Einstein College of Medicine.

Charles Weijer, MD, PhD, FRCPC, is Associate Professor of Bioethics and Surgery, Adjunct Professor of Philosophy, and Canadian Institute of Health Research Investigator at Dalhousie University in Halifax, Canada. He is currently Visiting Scholar in the Department of History and Philosophy of Science at the University of Cambridge and Visiting Fellow at Clare Hall in Cambridge, UK.

Simon Noble: There has been a lot of debate over the level of healthcare that AIDS vaccine trial researchers should be obligated to provide to trial participants who incidentally become infected with HIV during the course of the trial. Do you consider that there is a moral imperative for the trial researchers to provide this care, or is it simply morally praiseworthy to do so?

Ruth Macklin: I believe there is a moral obligation to provide antiretroviral treatment to trial participants who become infected with HIV during the course of the trial. However, to say that a moral obligation exists is not to define precisely on whom that obligation falls. Quite clearly, it is beyond the ability of a research team alone to provide ARV treatment. It is also likely that many sponsors will be unwilling to do so. Nevertheless, the obligation to provide treatment derives from considerations of justice. Trial participants who become infected deserve something in return for their contribution to vaccine research for which they have volunteered.

Charles Weijer: HIV vaccine research in developing countries is morally vexing for the same reason that it is of great social importance. It is carried out in the face of global disparities in the distribution of health care resources. Many people with HIV in developing countries simply don’t have access to life-saving HIV treatments. At the same time, the development of a successful HIV vaccine is one of the best hopes for lessening the burden of the disease in poor countries. Against this backdrop, the pressing nature of the question  ‘What do we owe participants in HIV vaccine trials conducted in developing countries?’ is immediate. My answer to the question is this: We owe them no more or less than is owed participants in HIV vaccine trials in developed countries. It is well accepted that research participation ought not disadvantage the medical care of patients (e.g., Declaration of Helsinki 28). Thus, HIV vaccine trials may only employ a placebo control so long as there is no effective HIV vaccine. Furthermore, should a safe and effective HIV vaccine emerge from a trial, it must be provided to all trial participants (Declaration of Helsinki 30). But, the claim that researchers have an obligation to provide treatment for those who develop HIV while enrolled in an HIV vaccine trial is both unprecedented and dubious. In moral theory, causation is a necessary condition of compensatory claims. Unless trial participation per se can be shown to have caused HIV infection (e.g., through the administration of an attenuated live vaccine), a compensatory claim upon the researcher is without merit.

Charles mentions precedent, but unlike previous vaccine trials for other diseases part of the success of an AIDS vaccine trial will be to measure secondary endpoints of vaccination, such as attenuation of disease. Aren’t AIDS vaccine researchers essentially studying disease course in those incidentally infected, and doesn’t this increase their obligation to these volunteers?

Ruth Macklin: The first thing to say about “precedent” is that it often does not provide a useful guide for what should be done in the future. If the world had to rely on precedents in determining what is the ethically right thing to do, there would be no moral progress. The pharmaceutical industry’s recent steps to reduce the price of AIDS drugs was also unprecedented, as are other current efforts to make ARVs widely available in developing countries. As for the question of whether researchers have an increased obligation when they study disease course in volunteers who become infected, some might say “no,” just so long as the volunteers are clearly informed at the outset that treatment will not be provided to anyone who becomes infected during the trial. There appears to be no more compelling reason to provide treatment when disease course is being studied in vaccine participants than in the case of epidemiologic studies of, say, disease transmission in discordant couples. The obligation to provide treatment to volunteers in AIDS research derives primarily from the principle of reciprocity: those who contribute to the research enterprise deserve something in return. Secondarily, there is a general ethical principle that mandates maximizing health benefits when possible. Taken together, these two principles give rise to the obligation to provide treatment to infected vaccine participants.

Charles Weijer: I agree that studying the course of disease does not provide a more compelling reason to provide treatment. The question says I “mention precedent.” I do not. I say that the claim that researchers have an obligation to provide treatment to those who develop HIV is unprecedented. This is simply a fact. An inference is criticized that I do not and would not make: future behavior should be guided solely by behavior deemed moral in the past. This view would put ethicists out of business and, frankly, I have a mortgage to pay. The claim I am making is this: unprecedented moral claims need to be supported by successful arguments before they are adopted as novel norms. Two arguments are provided in brief. Principles for the ethics of research are laid out in the Belmont Report (1979) and they are respect for persons, beneficence, and justice. There is no principle of reciprocity nor is there a principle of maximizing health benefits in the Belmont Report (beneficence requires that we maximize potential benefits to subjects, while minimizing risks—a different claim). Accordingly both arguments provided presuppose arguments not provided about the appropriateness of these novel principles for clinical research. Even the arguments provided lead to problems. Surely all research subjects in a clinical trial contribute to the success of the trial. If reciprocity means giving back to all those who contributed to the successful outcome, then providing a benefit only to a subset of participants violates reciprocity. A principle of maximizing health benefits leads to an unstoppable chain of claims against researchers unless “where possible” is precisely operationalized. Providing HIV treatment maximizes health benefits, but not as much as doing that and providing health care to everyone, and not as much as doing both of these and building hospitals too, and so on.

What is the source of ethical principles? Is the Belmont Report the only source, or the best one?

Ruth Macklin: Why should we consider the Belmont Report the last word in ethical principles—especially in the area of justice? The Belmont Report does not mention the principle proposed by John Rawls, that justice requires social arrangements to benefit the least advantaged members of society. Surely Rawls’s principle is a leading candidate among principles of distributive justice, even though it is not mentioned in the Belmont Report. Justice as reciprocity is simply one principle among many that are candidates for a just distribution of resources. It is certainly true that all research subjects contribute to the success of a trial. However, those who become infected contribute in a unique way. They are the only subjects whose contribution makes it possible to draw any precise conclusions about the efficacy (or lack thereof) of a preventive vaccine trial. And of course, a principle of health maximization is not intended to lead to an unstoppable chain of claims. Every principle requires interpretation and specifying criteria for its correct—and often limited—application. A full account of what is owed to research subjects and to others in the community or country would have to spell out such criteria and limits.

Charles Weijer: The Belmont Report is not the last word in ethical principles in research, but it is the first word. I take it as rudimentary that if one principle (e.g., reciprocity) implies a moral conclusion (e.g., provide treatment to those who become HIV infected in a vaccine trial) and another principle (e.g., Belmont principle of justice) implies the opposite, there are no rational grounds for the adoption of the first moral conclusion. A sound and convincing argument must be mounted that the principle of reciprocity is to be preferred over competing principles that imply opposite conclusions. Beyond this, I really don’t understand the claim that “those who become infected contribute in a unique way. They are the only subjects whose contribution makes it possible to draw any precise conclusions about the efficacy…of a preventive vaccine trial.” An HIV vaccine is effective if the proportion of subjects who receive the vaccine and become HIV infected is less than the proportion of subjects who do not receive the vaccine and become HIV infected. This is logically equivalent to saying that a vaccine is effective if the proportion of subjects who receive the vaccine and do not become HIV infected is greater than the proportion of subjects whodo not receive the vaccine and do not become HIV infected. Thus, by logic, the determination of an HIV vaccine’s efficacy relies no more on those who become HIV infected in the study than on those who do not. Per above, an argument is still required to demonstrate that a principle of maximizing benefit is to be preferred to the understanding of the principle of beneficence found in the Belmont Report. Beneficence has been specified in considerable detail (see: Journal of Law, Medicine & Ethics 2000; 28: 344-361) and this specification does not support any special obligations of researchers to research subjects in developing countries. We will have to await an equally careful specification—be it global or country by country—of the novel principle of maximizing benefit to see if it does avoid pitfalls and indeed whether it merits adoption at all.

One of the biggest fears among AIDS vaccine researchers is that an obligation to provide funds for treatment of incidental HIV infection will divert research funds, slowing the pace of research and ultimately delay the discovery of an effective vaccine.  With an estimated 14,000 new infections daily, any delay in the discovery process has profound ramifications.  Doesn’t this mean that the moral obligation should be toward the pace of discovery rather than treatment of infections that are not actually a research harm?

Ruth Macklin: The two goals—providing ARV treatment to infected research participants, on the one hand, and proceeding with all due speed in the research endeavor, on the other—are not incompatible. First of all there is no evidence whatsoever that research funds will be diverted to pay for treatment. Today there are numerous possible sources of funds and newly forged collaborative relationships among researchers, industry, private philanthropic organizations, and the public sector. There is a sense in which this entire discussion is dated. The World Health Organization (WHO) has recently announced its “3 x 5” initiative—to ensure that ARVs are available to 3 million infected individuals in developing countries by the year 2005. Surely, individuals who became infected during preventive vaccine trials can be among that group, as it contains many fewer than 3 million people.

Charles Weijer: Global initiatives, such as WHO’s “3 x 5” initiative, that seek to enhance the availability of ARVs to persons in developing countries further the cause of social justice and merit our support. The question in this exchange is whether it is prudent to saddle the research enterprise with this agenda of social reform. In its 1976 report on Research Involving Prisoners, the National Commission [for the Protection of Human Subjects of Biomedical and Behavioral Research] tied its recommendations to protect prisoners in research to an agenda of prison reform. Ostensibly to protect prisoners in research, the National Commission mandated 17 reforms to improve the living standards of prisoners, a suggestion greeted with enthusiasm. One ethicist, Nancy Dubler, who opposed research involving prisoners, commented at the time: “[S]imply imagine the strength of a lobby that would unite medicine, correctional officers, and drug companies. It would be in the overwhelming interest of all of these groups to continue a system that would provide an unlimited supply of available, trackable, and willing subjects.” The end result of this initiative is well known. Research in prisons came to an end and conditions in prisons continued to deteriorate. HIV vaccine research offers perhaps the best hope of ameliorating suffering from HIV in developing countries. Tying such research to an agenda of social reform runs the risk of slowing the pace of discovery or, worse yet, bringing it to a halt. 

What do you consider are the obligations towards potential trial participants who are found to be HIV infected at the initial screening (and are therefore excluded from the trial)? Should those excluded be offered ARVs and, if so, is that not an undue inducement to volunteer for trial participation in the first place?

Ruth Macklin: As the WHO’s “3 x 5” initiative is rolled out, there will have to be a careful analysis of fair principles for allocation of ARVs. It will be impossible to provide access to the entire 3 million initially. Beginning with groups that have already been tested and found to be positive might be one good place to start, from the standpoint of efficiency. Therefore, those found to be infected at the initial screening, as well as those who become infected during a trial, could be among the first groups since the VCT process is already in place. As for the “undue inducement” argument, it is a red herring. One worries about “undue inducement” when research carries high risks in an activity in which people would otherwise not choose to participate. But the current and proposed vaccine candidates are not risky. And people are eager to participate in the hope of being protected against infection. If anything could count as an undue inducement, it would be the vaccine itself, since that is what is hoped to provide the real benefit to uninfected people. But surely, we would not want to consider a preventive HIV vaccine an undue inducement, since in that case it would be unethical to conduct such research in the first place! 

Charles Weijer: The conflation of an agenda of social reform and obligations to research subjects is clear in this response. Individuals found to be HIV positive at screening are ineligible for study participation and are not research subjects. Putative principles of reciprocity and maximizing research benefit thus do not support an obligation on the part of researchers to provide such individuals with ARV treatment. In the context of a community in which ARVs are not available, their provision to persons found to be HIV positive at screening or to trial participants who develop HIV during the course of study is obviously undue inducement to participate in the study. Only study participation gives access to ARVs that can forestall what is otherwise certain death from HIV. Under these circumstances, the voluntariness of decisions regarding trial enrollment would be enhanced by either (a) not offering ARVs to persons found to be HIV positive at screening or to trial participants who develop HIV, or (b) initiating a community HIV testing and treatment program in parallel with the trial. If wide-scale ARV treatment in a particular community is not available through programs such as WHO’s “3 x 5” initiative, then the first option is preferred; if it is available, then the second option is preferred. Both options have the merit of mitigating undue influence and barring the injustice of treating trial participants preferentially to their neighbors not in the trial, whose right to treatment is equally pressing.

*Emily Bass is senior writer of the IAVI Report