Kimberly Page-Shafer [a], Vonthanak Saphonn [b], Ly Penh Sun [b], Mean Chhi Vun [b], David A Cooper [c] and John M Kaldor [c]

HIV prevention research in a resource-limited setting the experience of planning a trial in Cambodia

On Aug 13, 2004, all work was stopped and never resumed on a study being prepared in Cambodia, to determine whether daily oral tenofovir disoproxil fumarate was safe and effective in preventing the sexual transmission of HIV infection. The study was to have been a randomised, double-blind trial that recruited 960 HIV-negative women engaged in sex work to receive either 300 mg of tenofovir disoproxil fumarate once daily or an identical placebo, for 12 months. Trial development was suspended in response to a statement from the Cambodian Prime Minister.^[1Ð3] Before the statement, concerns about the trial had been raised by a group representing women engaged in sex work in Cambodia.^[4][5]

There has since been much commentary about the trial and its surrounding issues through the media^[6Ð9] and the internet. The commentary asserted that: the trial was purely motivated by the commercial interests of a pharmaceutical company, the drug had not been adequately assessed before it was proposed as a possible prevention agent, affected communities were inadequately consulted during study preparations, the researchers would not be providing participants with safe sex counselling and condoms during the trial to increase HIV incidence, recruitment practices would not adequately incorporate informed consent, and medical treatment for trial-related injury and HIV care for seroconverters would not be available to participants. Along with this commentary, the researchers were accused of engaging or intending to engage in practices that were inconsistent with the standards of research ethics. To better inform public discussion, we provide our perspective on the design and development stages of the trial, focusing particularly on issues raised by commentators.

Trial researchers, sponsors, rationale, and setting

The study was planned as a collaboration between the University of California, San Francisco, CA, USA (UCSF); the University of New South Wales, Sydney, Australia (UNSW); and the National Centre for HIV/AIDS, Dermatology and Sexually Transmitted Infections, Phnom Penh, Cambodia (NCHADS). The funding sponsors were the US National Institutes of Health (NIH) and Family Health International, a US-based research organisation that had received a grant from the Bill and Melinda Gates Foundation to investigate tenofovir disoproxil fumarate as an HIV prevention drug at sites in Africa and Asia. Gilead Sciences, the drug manufacturer, provided the drug and placebo, but had no role in the conduct of the study.

HIV prevention through the promotion of behaviour change, including increased condom use, has had some success in reducing transmission rates, although incidence has remained intractably high in many parts of the world. Women are especially vulnerable, since they are not always in a position to negotiate safer sex practices.

It is widely recognised that biomedical interventions will be needed in addition to effective behavioural interventions, but effective prophylactic vaccines or microbicides are probably many years away, despite intensive research efforts. Pre-exposure chemoprophylaxis involves the use of antiretroviral drugs before exposure to HIV. The potential of this approach is shown by the use of antiretroviral drugs for post-exposure prophylaxis and in the prevention of mother-to-child transmission.^[10][11] Tenofovir disoproxil fumarate, an anti-HIV drug that works by blocking reverse transcriptase, is a promising candidate for chemoprophylaxis because of its once-daily dosing schedule, its low levels of side-effects seen in major clinical studies and in several hundred thousand people with HIV infection who have been treated with the drug, and its protection against simian immunodeficiency virus (SIV) transmission in animal models.^[12Ð15] Although the emergence of resistance is a potential risk of chemoprophylaxis, the main resistance-conferring mutation of tenofovir disoproxil fumarate seems to develop relatively slowly,^[15] which favours its use from this perspective.

The trial was proposed for female sex workers in Cambodia because comprehensive surveillance has shown that HIV incidence and prevalence in the country are among the highest in the Asia-Pacific region. HIV prevention, mainly based on condom promotion, has been implemented on a large scale, and in the past 3Ð4 years, the country's comprehensive sentinel surveillance programme has detected falling HIV prevalence in most population groups including sex workers.^[16] Nevertheless, sex workers, both brothel-based and non-brothel based, still remain at high risk of infection,[16] indicating a continuing need for female-controlled HIV prevention technologies.

Although tenofovir disoproxil fumarate is being tested as a prevention drug in several countries, including trials of women at increased risk in Africa,^[17] the results of these investigations might not be directly relevant to women in Asia, because of differences in overall health status, body size, attitudes to medical intervention, and possibly genetic factors.^[18][19] The prevention strategy would need to be assessed in a similar population at risk of HIV infection, and ideally one that would stand to benefit should the strategy prove effective.

Consultation and approval processes

Research of this kind involves multiple levels of approval. First, we required government approval before we could undertake any form of trial-related activity, including community consultation, in Cambodia. An early version of the study protocol was approved by the Cambodian Ministry of Health National Ethics Committee (NEC) in July, 2003. Approvals of this early version were obtained from the ethics committees at UCSF in March, 2003, and at UNSW in August, 2003. Contracts to develop the laboratory and clinical site and to undertake formative work with communities affected by the trial were negotiated between UCSF, UNSW, and NCHADS from January, 2003, and a long process of study development began. We understood from the outset that these initial approvals only provided authorisation to begin preparatory work, and that recruitment could not begin until the protocol had undergone all necessary revisions and been approved in final form, by the Cambodian NEC, and ethics committees representing the US and Australian collaborators.

Second, as the main trial sponsor, the NIH extensively reviewed the study plans during the first half of 2004, including a series of inspection visits to Cambodia to assess the laboratory, pharmacy, clinical site, data management, and general level of community and institutional preparation. As a result of this review and ongoing community input (which we discuss later), the study went through several revisions, and in July, 2004, the NIH declared support for resubmission to the NEC. The NIH had intended to take out an investigational new drug application through the US Food and Drug Administration, which would have given an extra level of protocol review and scrutiny over the conduct of the trial.

A third level of review was via the advisory structures established for the trial in Cambodia. The Cambodian Ministry of Health established an external advisory board, which met in January and June, 2004, and brought together key governmental departments and international organisations with an interest in HIV and AIDS, including UNAIDS (Joint UN Programme on HIV/AIDS) and WHO. NCHADS assembled a community advisory group that met in March and May, 2004, to provide a more direct voice for non-governmental organisations and sex workers in particular. The concept and processes of a community advisory group were new to both researchers and stakeholders in Cambodia, and procedures were being developed with international guidelines. Having each had the opportunity to hold only two meetings by the time the trial was suspended, neither the external advisory board nor the community advisory group could establish efficient operating procedures. Despite intentions, the community advisory group had little success in facilitating involvement of individual sex workers.

In addition to these formal structures, NCHADS held well-attended public meetings in July, 2003 and March, 2004, which generated lively debate across a range of topics. During this period, contacts were made with various community organisations, including sex worker networks, to foster dialogue on the trial. From February to August, 2004, after which continuing trial preparations were stopped, researchers at NCHADS did formative research with sex workers, brothel owners, and other key stakeholders such as community leaders and the police. This research included a series of focus groups and key informant interviews that were to guide the development of the main aspects of the study, particularly the process of informed consent, and recruitment and retention procedures. Issues discussed during these meetings included the rationale and design of the trial, ways in which participants might wish to be contacted, and the amount of money that would be appropriate for reimbursement of study visits. Study materials were developed in the Khmer language (eg, a protocol summary, informed consent form, and illustrated booklet), and participants in focus groups were asked to review and comment on these materials. ^[20] such sessions were held, and the results were steadily incorporated into the protocol development, again resulting in revisions to the proposed trial operations.

This detailed description is by no means intended to try to prove that the researchers established an ideal consultation process, and indeed we clearly had much to learn from the outset. We were well aware that the very idea of a prevention trial was new in Cambodia, and that there was little experience of collaboration between researchers and community. Despite the establishment of these consultative mechanisms and our commitment to continue dialogue for the development of the study plans, we recognise that not all those who engaged with us felt a genuine sense of involvement — some individuals and organisations chose not to participate in meetings or community forums. We also perceived that the constraint on US-funded agencies with respect to sex-worker-related projects might have contributed to a sense of disfranchisement among some groups. When concerns about the trial were first raised publicly,^[4][5] it became clear to us that aspects of the trial plans were being portrayed inaccurately. At this point, we proposed to our funding agencies that the then current protocol be posted on our institutional websites, as a means of publicly presenting the trial status and planning, but this proposal was rejected. In any case, despite our efforts at communication, key issues of concern remained in contention when the trial-related activities were stopped. We review the problems that presented substantial challenges to the design and conduct of the study.

HIV prevention in trial participants

The best standard of care for HIV prevention in individuals that is currently available is the provision of information about HIV transmission and how infection can be avoided, condoms to people who may be at risk through sexual exposure, and sterile injecting equipment to people at risk through injection practices. Participants in the planned prevention study in Cambodia were to receive safe-sex counselling and condoms at every monthly visit during their year's enrolment in the trial. Since injecting drug use is thought to be very rare in women engaged in sex work in Cambodia, there were no plans to distribute sterile injecting equipment.

The informed consent process for the trial was to have emphasised that participants could not assume that they had any additional protection against HIV infection, for two reasons. First, whether tenofovir disoproxil fumarate will actually prevent infection is not known, and second, half the participants would not actually receive the drug. Participants' risk behaviour would have been monitored throughout the trial and counselling procedures modified if increased risk behaviour were detected. Experience from vaccine research suggests that most trial participants do reduce their levels of risk behaviour, although these changes vary by group and over time.^[20][21]

Provision of non-HIV-related health care to trial participants

With its extensive use to treat people with HIV infection, tenofovir disoproxil fumarate is recognised as having a very low side-effect profile compared with other antiretroviral drugs. However, data are scarce for the effects of tenofovir disoproxil fumarate in HIV-negative people and acceptability may be lower in these individuals than for those receiving therapeutic benefit. Therefore, a comprehensive series of biochemical, haematological, and clinical safety assessments was scheduled on a monthly basis.

Study participants who had any adverse health event during the trial were to receive free clinical care when necessary, either at the trial site or from carefully selected medical specialists. As with other clinical trials, an independent data safety monitoring board, with representation from Cambodia as well as from the NIH, would have regularly monitored the safety data accumulated during the course of the trial, and had the authority to stop the trial on safety grounds.

The post-trial provision of care became one of the key points of debate surrounding the trial. In countries with well-established accessible health care and social-security systems, people who have long-term illness or injury due to their participation in a drug trial can expect to have their health needs met, receive some form of income replacement, and even have legal recourse to apply for compensation through the court system. In countries with limited resources, such as Cambodia, these options are not generally available, and it is understandable that participants with very low income would want to receive some form of provision for their future clinical care and income support, should they be injured through the trial.^[22][23] Such coverage is in fact required by national ethical guidelines in many countries, including Cambodia.^[24]

This issue was of concern to us from an early stage. During 2003, we consulted with investigators who had faced similar issues in other developing countries, and considered various options for the assistance of people with long-term health problems that could be attributed to their participation in trials, including insurance schemes, lump-sum payments, and the establishment of long-term contracts for the provision of clinical services. None of these options fits with the policies of our funding agencies, and despite the requirements of Cambodian guidelines, we found ourselves unable to enunciate a clear response to the concerns of organisations that spoke on behalf of potential participants. We continued to engage with our funding agencies about this issue, and achieved some resolution, in the form of a commitment to fund medical care during the trial, but not afterwards. One solution that remained open to us was the identification of alternative sources of funding, to provide long-term coverage for trial-related harm.

Provision of HIV-related health care to trial participants

The occurrence of HIV infection in prevention trial participants was a primary endpoint of the trial, as well as being a serious adverse event. There would seem to be both a medical and ethical requirement for the study investigators to provide appropriate care to any participant who acquired HIV infection during the course of the trial, including antiretroviral drugs as clinically indicated.^[22][25] As with non-HIV-related health care, sustained provision of medical care after the trial would be very difficult, when the infrastructure and funding mechanism responsible for the study would no longer be present.

We made arrangements for participants who acquired HIV infection during the trial to be referred to the NCHADS HIV clinic, which was then under development and opened in November, 2004. The clinic provides free medical care through funding that is independent of the resources made available for the HIV prevention trial. NCHADS agreed to guarantee access for seroconverting participants to the clinic for assessment and care, according to Cambodian national treatment guidelines,^[26] which are based on WHO guidelines for antiretroviral therapy in resource-limited settings.^[27] The long-term availability of antiretroviral drugs can never be taken for granted, but NCHADS's success in gaining funding for the large-scale supply of antiretroviral drugs to the end of 2009 through the fourth round of the Global Fund was a promising sign for future treatment. The NCHADS HIV clinic has now seen more than 400 people with HIV infection, of whom over 150 have begun antiretroviral therapy.

Potential participants who have HIV infection at screening would have been referred for treatment and care to the NCHADS HIV clinic or other HIV treatment services operating in Phnom Penh. Study investigators must ensure that HIV testing for potential trial participants is undertaken with a meticulously administered informed consent explaining the risks and benefits of testing and accompanied by comprehensive pre-test and post-test counselling. However, we would not have guaranteed a preferential referral for those who tested positive at screening, since the bypass of existing selection procedures for antiretroviral drug programmes through research screening would probably be perceived as disadvantageous to people accessing HIV care by established mechanisms.

What remains?

At the point when the trial planning was stopped, a new clinical facility had been built, the national laboratory had been supported to develop its procedures to an international standard of quality, and more than 40 local staff were working on study preparations. The laboratory remains functional, serving Cambodia for clinical and public health purposes, and several key staff were employed on other research projects.

Funding for the prevention trial was sought and obtained, on the premise that it would be important to know whether tenofovir disoproxil fumarate was an effective prevention agent. The trial's success depended on an assumption that there would be a substantial number of women engaged in sex work in Cambodia with a sufficient interest in determining the answer to the question to want to participate. From this experience, we know that there were important concerns among potential participants about how such a trial might be undertaken, but whether these concerns could have been resolved to the satisfaction of most remains unknown. We are acutely aware that once doubts were raised, those who were speaking on behalf of the participants thought there was no forum in which they could engage with the researchers to bring about meaningful changes in the study design, processes, or conduct, despite our various attempts to provide such channels and to indicate that such changes were possible.

Although we remain confident that our trial planning followed, and in many respects went well beyond, accepted ethical standards, we have absolutely no doubt that, as with any human endeavour, there were areas in which the management of interactions with potential participants and their representatives could have been substantially improved. It would be simplistic for us to look back on this experience and state that this process or that strategy could have led to different outcomes, but it is nonetheless instructive to consider the different pathways that could have been taken.

Because prevention research requires work at a community level, trust can never be in short supply, and the key to trust is communication. Establishment of new research programmes across cultures demands special efforts and innovation in communication, and there is little in the way of textbooks and guidelines to show the way to achieve these goals. Although it is an unusual circumstance for a national leader to take a direct interest in the conduct of a trial, the involvement of the Cambodian Prime Minister showed a genuine concern for issues that remain challenging, not only in Cambodia but wherever a prevention trial is being considered in a population with limited access to basic health services. Communities, governments, researchers, and funding agencies should come together for frank discussions about these issues, if we are to be able to test critical approaches to prevent the transmission of what is still regarded as the greatest global threat to human health.

Conflict of interest statement

D A Cooper is a member of the Gilead International and Australian Advisory Boards, and has been a speaker at Gilead-sponsored symposiums. The other authors declare that they have no conflict of interest.

Acknowledgments

This study was funded by the US National Institute for Allergy and Infectious Diseases Division of AIDS (5UO154241), and by a subcontract with Family Health International with funding from the Bill and Melinda Gates Foundation. The US National Institute for Allergy and Infectious Diseases Division of AIDS participated in study protocol review and approval. Funding agencies did not participate in writing the report. The drug manufacturer, Gilead Sciences, were to provide tenofovir disoproxil fumarate and placebo free of charge, but provided no further funding and had no role in the conduct of the study or preparation of the report. We thank key study personnel (Shannon Casey, Julian Elliott, Jennifer Evans, Robert Grant, Iona Millwood, Robert Oelrichs, Maria Lucia Pecoraro, Phlong Pisith, Mary Poynten, Jenne Roberts, Soleak Sim, Chuop Sokheng, Ellen Stein, Khol Vohith) who were dedicated to the development of this trial for 2 years; the US National Institute for Allergy and Infectious Diseases Division of AIDS; Family Health International; Gilead Sciences and the Bill and Melinda Gates Foundation; members of the External Advisory Board and the Community Advisory Board; and all participants in community forums, focus groups, and in-depth interviews that were held as part of trial development. The National Centre in HIV Epidemiology and Clinical Research is funded by the Australian Government.

References

Affiliations

1. Center for AIDS Prevention Studies, Department of Medicine, University of California, San Francisco, CA, USA
2. National Centre for HIV/AIDS, Dermatology and STDs, Ministry of Health, Phnom Penh, Cambodia
3. National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Darlinghurst, Australia

Correspondence to: Prof John Kaldor, National Centre in HIV Epidemiology and Clinical Research, Darlinghurst, New South Wales 2010, Australia

[Reference: The Lancet 2005; 366:1499-1503, DOI:10.1016/S0140-6736(05)67146-2