Article Text

Download PDFPDF

First-in-human HIV-remission studies: reducing and justifying risk
  1. Rebecca Dresser
  1. Correspondence to Professor Rebecca Dresser, Washington University Law School, One Brookings Drive, Box 1120, St. Louis, MO 63130, USA; dresser{at}


Interest and excitement surround the possibility of developing measures that produce sustained or permanent HIV remission in infected individuals. First-in-human (FIH) trials are one step in exploring this possibility. Initial human trials raise the usual ethical issues associated with human research, and a set of distinct issues. Because the potential direct benefits to FIH trial volunteers will be either small or non-existent, trial risks must be justified by the social value of the information the trials are expected to produce. To minimise and justify risks, researchers must have solid preclinical evidence that FIH trials will be safe and produce information relevant to human health improvements. Researchers must also adopt adequate study safeguards to protect FIH subjects. An ethically defensible study population must be selected as well.

  • Research Ethics
  • HIV Infection and AIDS

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Scientists are working towards a new goal in HIV research. The Berlin patient's apparent cure, together with other cases in which infected individuals remained virus-free for lengthy periods, suggest that it may be possible to develop interventions that cure or, alternatively, produce long-term antiretroviral-free remission.

First-in-human (FIH) trials are one step in exploring this possibility. Many such trials are under way and many more will be proposed in the coming years. Researchers conducting early human trials are investigating a variety of strategies, including intensified drug regimens, potential therapeutic vaccines, immunotherapy, monoclonal antibodies, gene transfer agents and stem cell transplantation.1

In this paper, I describe how previous experience with FIH trials points to ways of minimising and justifying risks to subjects in early phase HIV-remission trials. (I use the term HIV remission because it more accurately describes the broad range of studies that are typically classified as HIV-cure trials.) Initial human trials raise the usual ethical issues associated with human research, and a set of distinct issues. These issues are related to the relatively high uncertainty that exists when the first humans are exposed to an investigational intervention and to the fact that FIH studies rarely provide direct health benefits to subjects.

Background considerations

Ethical principles and regulatory rules set standards for FIH studies. Risks to human research subjects must be minimised and justified by the value of the knowledge an initial human study is expected to produce. Some ethics statements also set limits on permissible risk to human subjects. For example, according to the Nuremberg Code, ‘No experiment should be conducted where there is an a priori reason to believe that death or disabling injury will occur’.2

The Nuremberg Code specifically addresses FIH studies, as well, describing the evidentiary foundation that should underlie such studies: “The experiment should be so designed and based on the results of animal experimentation and knowledge of the natural history of the disease or other problem under study that the anticipated results will justify the performance of the experiment”.2

Initial human studies are designed to evaluate the human body's response to an investigational intervention. Most familiar are phase 1 trials evaluating novel investigational drugs. According to US Food and Drug Administration regulations governing drug approval, phase 1 studies are ‘designed to determine the metabolism and pharmacologic actions of the drug in humans, the side effects associated with increasing doses, and if possible, to gain early evidence on effectiveness’.3 Phase 1 and other FIH studies do not necessarily present high risks to subjects; rather, as I discuss below, the magnitude of risk to FIH trial subjects varies depending on specific trial characteristics.

Uncertainty is one of the major factors that distinguish FIH trial risks from risks presented in later-phase trials. Because the predictive value of animal and other preclinical research is limited, human subjects in FIH trials may respond in unanticipated ways. This feature of FIH trials underlies many of the specific ethical considerations that apply to such trials.

The lack of direct health benefit to subjects is the second major factor affecting FIH trial ethics. Subjects in FIH trials rarely experience health improvements from trial interventions; indeed, most FIH trials are not designed to produce such effects. Instead, FIH trials are designed to evaluate safety and to generate information to guide decisions about future investigations.4

Some people participate in FIH research because they expect to receive psychological benefits from contributing to the research effort. People also enter FIH trials because they believe they might personally benefit from extra examinations, tests and other procedures included in the trials. Although the desire to secure such collateral benefits motivates some individuals to become FIH study participants, trials offering collateral benefits to subjects must also offer the knowledge advances that make research worthwhile from a societal perspective.5

Risks to FIH trial volunteers must be justified by the social value of the information such trials are expected to produce.6 ,7 The social value of study information is a function of the level of improvement the hoped-for intervention will provide to patients.8 Although current therapies enable many HIV-infected persons to survive and thrive, existing therapies also produce side effects and long-term complications. Moreover, current therapies are expensive and don't work well for everyone. Thus, finding more effective and less costly HIV therapies would have high social value.

Determining the value of specific FIH HIV-remission trials involves two additional steps, however. These steps address the scientific quality of the trials. To have value, trials must be designed to produce valid data that will answer the questions the trials are investigating. Trials must also generate knowledge that will help determine whether an intervention is sufficiently promising to warrant further investigation.8 These requirements are essential to conducting ethical FIH studies because studies that fail to produce valuable knowledge expose study subjects to risk for no good reason. Accordingly, researchers should seek to minimise risks to subjects, and to maximise the value of FIH HIV-remission trials.

In sum, before permitting people to enrol in FIH trials, investigators, sponsors, regulators and ethics review committees must find that the trials present an acceptable balance of risks and anticipated advances in valuable knowledge. Conventional research ethics and regulatory requirements incorporate a paternalistic constraint on the individual's choice to enrol in research. This constraint is justified because prospective subjects lack the expert knowledge that would enable them to independently protect their personal interests in the research context. Another justification for the constraint is that many people with serious medical conditions are unduly optimistic about the personal health improvements available through FIH study participation.9

Preclinical evidence and FIH trials

Preclinical investigation (in vitro and animal research) plays a major role in applying ethical principles to FIH trials. Before beginning an FIH trial, scientists must have an adequate theoretical and empirical basis for proceeding. Otherwise, human subjects will be exposed to unnecessary and unjustified risks. Limited financial and other research resources will be wasted, too.10

When FIH trials go forward without an adequate preclinical basis, unexpected risks to subjects can materialise. One source of this problem is the use of an inappropriate animal model. The use of an animal model that fails to mimic relevant human effects exposes FIH trial subjects to unexpected harm. For example, unexpected harm materialised in the 2006 FIH trial involving a novel monoclonal antibody called TGN1412. Human subjects in that trial suffered serious, life-threatening adverse events. This was a surprise because researchers had previously detected no adverse effects in monkeys given a much higher dose of the antibody than the human subjects received. Experts who later reviewed the trial said that the reliance on monkeys could have been misplaced, given differences in the relevant monoclonal antibody receptors in human and non-human primates.11

Besides failing to predict human risks, inadequate preclinical research can predict beneficial effects that fail to materialise in humans. Once again, inappropriate animal models can contribute to this problem. Stroke research is often cited as an area in which many interventions that were effective in laboratory rodents produced no benefit in human trial subjects.12

Other deficiencies in preclinical research are linked to problematic methods and practices in laboratory animal research. The failure to engage in long-term monitoring of study animals means that long-term risks like tumour formation can go undetected.13 Animal study findings are often less valid than they could be because investigators fail to adapt a priori hypotheses and power calculations. Additional shortcomings in animal studies include lack of randomisation and blinding and failure to properly apply inclusion and exclusion criteria to study animals.14

Reporting biases may also exaggerate the significance of preclinical study results. The failure to publish negative results, as well as selective data analysis and outcome reporting, contribute to inflated claims about the significance of preclinical study findings. Greater transparency in publicising preclinical study results is also needed to improve decision-making about the effects an intervention could have on humans.15 These problems are often cited in discussions of translational science disappointments.14

Demanding high-quality preclinical research as the foundation for FIH HIV-remission trials would reduce risks to FIH subjects and increase the value of the information produced in human trials. The preclinical evidence supporting such trials should be carefully scrutinised. Investigators proposing FIH trials should offer substantial preclinical data indicating that interventions are reasonably likely to be both safe and effective in humans. Those data should undergo ‘rigorous and independent peer review’ to ensure that they supply an adequate basis for moving to initial human testing.16

Protective safeguards in FIH trials

Subjects in HIV-remission trials will face different degrees of risks and uncertainty. Some trials will present more predictable risks than others, based on their resemblance to previously evaluated interventions. For instance, one analysis describes three potential therapeutic approaches aimed at HIV-remission: gene-targeted therapy; chemotherapy; and stem cell therapy.17 Some risks presented by these three approaches are relatively well known, for applications aimed at other diseases have been extensively evaluated in human trials and used in patient care. But these and other investigational approaches to HIV remission will also involve new research populations and other novel features. In this situation, strict safeguards are needed to limit human exposure to unexpectedly toxic novel agents.

To protect FIH subjects, researchers must choose an appropriate starting dose and method of administering an investigational agent. Starting doses in FIH trials should take into account safety concerns and uncertainty, as well as size and other differences between human and non-human species. Human doses usually are based on the highest dose that produced no detectable adverse effects in the most relevant animal model. But experts support using a different approach when there is reason to suspect that humans could respond differently to an intervention than study animals did. In such cases, they recommend basing the human dose on the highest dose that produced no detectable effect of any sort in study animals.11

Problems arise when FIH trials incorporate measures that are substantially different from those used in the animal studies that preceded human trials. Substantial deviations from preclinical studies put human subjects at risk and undercut the rationale for conducting preclinical research. To preserve the integrity of the research enterprise, ‘delivery strategies, targets, doses, and materials in the human study [should be] identical or substantially equivalent to those validated in animals’.10

The risk reduction principle also requires investigators to provide an adequately trained study staff and adequate medical facilities to ensure that subjects receive appropriate care for any complications that arise in FIH studies. Trials involving interventions with potential serious acute effects should include an observation period between each subject's exposure. Because the TGN1412 antibody trial failed to include such an observation period, more subjects than necessary suffered harm from the trial intervention.9

Well founded and clear stopping rules are additional measures that can protect subjects in FIH HIV-remission trials. Study planners and reviewers must also seek to minimise the risk that prior FIH study participation will disqualify patients from enrolling in later-phase HIV-remission trials or reduce their ability to benefit from future therapies.6 As in other kinds of human studies, strong confidentiality protections and injury compensation programmes can reduce harm to HIV-remission trial subjects, too.

Subject selection and risk minimisation

The choice of FIH study population is another element of risk minimisation. Healthy volunteers are the population of choice in most FIH trials. But a different approach is applied to certain FIH trials. In some situations, people with the condition of interest are recruited for FIH research.

The choice of subject population depends partly on a trial's scientific objectives. Sometimes the best data will come from people in good health with relatively low previous exposure to medications and other agents. Sometimes, however, the most informative data will come from people with the health problem the investigational intervention is targeting.18

Ethical considerations influence the choice of study population, too. This dimension of the decision turns on the relative risk that subjects will face in a trial. The conventional ethical judgement is that higher risks are acceptable in FIH trials involving people who already face lethal and other serious risks from a medical condition. On this analysis, healthy individuals who die or experience other serious harm as a result of research participation experience a greater loss than do participants facing death or serious harm from an untreatable condition. Accordingly, research risks that would be unreasonable for healthy people to assume become reasonable risks for people already facing death and other serious harm from pre-existing disease or injury. These judgements account for the long-standing practice of relying on patients with untreatable disease to serve as subjects in FIH chemotherapy trials, as well as in initial human trials of other risky interventions, such as artificial heart devices.19

The practice of relying on patients with untreatable conditions for FIH trials remains somewhat controversial, however, for two reasons. One is that even patients with untreatable life-threatening conditions can experience serious losses in early phase trials. Such patients can have shorter and less comfortable lives due to complications and side effects they experience as FIH trial subjects.20

A second ethical concern complicates the choice of patients as FIH trial subjects. Some ethicists argue against recruiting patients who fail to benefit from existing therapies out of worry about the quality of such patients' choices. According to this view, patients' lack of good treatment options can make them desperate for therapeutic alternatives. This mindset can in turn make it difficult for them to appreciate that the decision to join an FIH trial has little or no chance of improving their medical situation.21

Members of another potential FIH study population known as stable patients have more attractive treatment options than do patients who fail to benefit from existing options. As a result, stable patients tend to be more psychologically equipped to understand that altruism, rather than hope of personal benefit, is the most realistic basis for enrolling in FIH trials. On the other hand, stable patients enrolling in FIH trials face higher relative risk than do patients with untreatable conditions. In these circumstances, the ethical principle of risk reduction conflicts with ethical principles promoting informed and voluntary choice to research participation.

Two examples illustrate the ethical controversy that can surround the choice between untreatable and stable patients as FIH study subjects. One example involved Jesse Gelsinger, a young man with a treatable genetic condition who died after being exposed to an FIH gene transfer trial intervention. Researchers had chosen to enrol stable patients in the gene transfer trial, rather than infants with a fatal form of the genetic condition under study.

The researchers' choice of study population was based partly on concern that the infants' parents would be quite desperate for therapeutic options. This desperation, it was argued, would keep many of them from recognising that their children had almost no chance of being helped by the trial intervention. Although some reviewers argued that terminally ill infants would be the more defensible study population, trial investigators decided that adult stable patients would be more suitable. As a result of that decision, stable patients like Gelsinger were exposed to relatively high risk in the FIH gene transfer trial. In the end, Gelsinger's death spurred public outcry and led to several years of questions and delay for the gene transfer research field.22 ,23

Another subject selection controversy surrounded the first human trial evaluating a human embryonic stem cell intervention. The trial sought to evaluate the intervention's effect in people with spinal cord injuries. Researchers in the stem cell trial planned to recruit patients who had recently suffered a complete spinal cord injury. But critics argued that recently injured patients would be too susceptible to the unrealistic hope that trial participation would improve their condition. Critics of the plan also argued that trial risks were excessive, for participation could jeopardise patients' 6–10% chance of spontaneous functional improvement. Critics contended that patients with chronic complete spinal cord injuries would be a more ethically defensible FIH study population, because they would face lower relative risk and be better equipped to understand the trial's risks and potential benefits.24

The stem cell investigators decided to stay with their original subject population and went forward with the trial. The company sponsoring the trial cancelled it after enrolling just four subjects, however, citing business considerations. The company's chief executive reported that researchers had seen no evidence of either safety concerns or health benefits for the subjects that had enrolled.25 But this information has limited value, given that the trial was terminated prematurely. Moreover, the choice of subject population remains controversial.

The choice of subject population for FIH HIV-remission studies is a ‘deep ethical question’.26 Should HIV-remission trials follow the lead of early phase chemotherapy trials and rely on subjects lacking acceptable therapeutic options? Or should HIV researchers adopt the stable patient model? The challenge for trial planners is to determine ‘the least vulnerable population that can usefully answer the research question’.24

Stable patients are likely to be the scientific population of choice for many FIH HIV-remission trials. In this research context, the stable patient population consists of people doing well on currently available antiretroviral medications. Individuals whose health status and quality of life are good with existing therapies would face high relative risk in FIH HIV-remission studies. In contrast, individuals who cannot be successfully treated with existing therapies would face lower relative risk in FIH trials. If HIV-remission researchers can obtain the information they need from FIH studies involving patients who fail to benefit from existing therapies, enrolling such subjects would be a risk-reduction measure. But patients with HIV lacking good therapeutic options could be susceptible to unrealistic optimism about the health improvements available through FIH trial participation.

Researchers studying HIV remission should carefully evaluate their scientific needs. If basic FIH information can be obtained from volunteers who have not done well on existing therapies, researchers should recruit from this population. This study population's lack of treatment alternatives does not diminish researchers' duties to minimise and justify risks, however.20 Before recruiting members of any population for FIH trial participation, there should be a ‘high standard of consensus in the scientific community that the FIH study is justified’.27 Furthermore, when recruiting any individual with HIV for FIH trials, researchers must ensure that study materials and discussions alert prospective subjects to the trial's knowledge-seeking purpose and the low likelihood of direct health benefit. Study teams should also adopt measures to assess and improve subjects' comprehension of these study facts.6 ,7

When FIH trials seek data that only stable patients can provide, researchers will seek to recruit those patients for trials. Because such patients face higher relative risk in FIH trials, adopting preclinical and FIH trial strategies to reduce risks to subjects will be extremely important. Although researchers should promote study understanding among all HIV trial subjects, they have a particularly strong obligation to ensure that stable patients understand the potential benefits and harms of FIH trial participation.28


Careful planning is essential to conducting FIH trials aimed at developing effective approaches to long-term HIV remission. Because HIV-remission studies are a relatively recent development, the research community has a good opportunity to establish strong ethical standards for the field.

Researchers proposing FIH HIV-remission trials have ethical responsibilities to minimise risk by demanding high-quality preclinical evidence to support initial human trials. In light of the uncertainty characterising FIH trials, researchers must build in safeguards to minimise risk as well. Trial planners must also consider the ethical advantages and disadvantages of relying on stable patients to serve as FIH trial subjects.

Premature and poorly designed FIH trials expose subjects to unnecessary and unjustified risk. When FIH trials go awry, the resulting outcry can also set back a research field.29 The low-quality data generated by premature and poorly designed studies can discourage scientists from pursuing what might actually be promising avenues of investigation.10 By learning from previous FIH trials, the HIV research community can effectively protect human subjects and advance scientific inquiry into potential strategies for inducing long-term HIV remission.


The author thanks Nir Eyal and other project members for helpful comments on an earlier draft of this article.



  • Funding US National Institute of Allergy and Infectious Diseases, R01 AI114617, ‘HIV Cure Studies: Risk, Risk Perception, and Ethics’.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Other content recommended for you