Europe PMC

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

Ethical tradeoffs in trial design: case study of an HPV vaccine trial in HIV-infected adolescent girls in lower income settings.

Author information

Affiliations

  • 1. Harvard School of Public Health – Center for Biostatistics in AIDS Research, 651 Huntington Avenue, Boston, MA 02115, USA.
      Authors
    • Lindsey JC 1
    (1 author)

ORCIDs linked to this article

Developing World Bioethics, 31 May 2013, 13(2):95-104
https://doi.org/10.1111/dewb.12028 PMID: 23725055 PMCID: PMC3716837

Free full text in Europe PMC

Abstract 

The Declaration of Helsinki and the Council of the International Organization of Medical Sciences provide guidance on standards of care and prevention in clinical trials. In the current and increasingly challenging research environment, the ethical status of a trial design depends not only on protection of participants, but also on social value, feasibility, and scientific validity. Using the example of a study assessing efficacy of a vaccine to prevent human papilloma virus in HIV-1 infected adolescent girls in low resource countries without access to the vaccine, we compare several trial designs which rank lower on some criteria and higher on others, giving rise to difficult trade-offs. This case demonstrates the need for developing more nuanced guidance documents to help researchers balance these often conflicting criteria.

Free full text 

Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev World Bioeth. Author manuscript; available in PMC 2014 Aug 1.
Published in final edited form as:
PMCID: PMC3716837
NIHMSID: NIHMS448599
PMID: 23725055

ETHICAL TRADEOFFS IN TRIAL DESIGN: CASE STUDY OF AN HPV VACCINE TRIAL IN HIV-INFECTED ADOLESCENT GIRLS IN LOWER INCOME SETTINGS

J. C. LINDSEY, S. K. SHAH, G. K. SIBERRY, P. JEAN-PHILIPPE, and M. J. LEVIN
Author information Copyright and License information Disclaimer

Abstract

The Declaration of Helsinki and the Council of the International Organization of Medical Sciences provide guidance on standards of care and prevention in clinical trials. In the current and increasingly challenging research environment, the ethical status of a trial design depends not only on protection of participants, but also on social value, feasibility, and scientific validity. Using the example of a study assessing efficacy of a vaccine to prevent human papilloma virus in HIV-1 infected adolescent girls in low resource countries without access to the vaccine, we compare several trial designs which rank lower on some criteria and higher on others, giving rise to difficult trade-offs. This case demonstrates the need for developing more nuanced guidance documents to help researchers balance these often conflicting criteria.

Keywords: standard of care, research ethics, public health, vaccines, clinical trials

INTRODUCTION

Two controversies have dominated ethical debate over standards of care in clinical trials. The first arose after a landmark placebo-controlled trial showed that an intensive regimen of azidothymidine (AZT) could reduce mother-to-child transmission of the human immunodeficiency virus (HIV) by 70%1. Given the high cost and complexity of the study regimen, it was not clear that it could be implemented in lower income countries. Follow-up studies were conducted using shorter, cheaper, and easier-to-administer regimens, which were ultimately found to significantly reduce HIV transmission to non-breastfed infants using an endpoint assessed six weeks after birth. All but one of these sixteen trials compared short course AZT to a placebo2. The second controversy involved a proposal for a placebo-controlled trial of a new surfactant (Surfaxin) administered to premature infants with underdeveloped lungs in Bolivia3. Surfactants were not available to infants in lower income countries and were unlikely to be made available in Bolivia after the study. Both cases were criticized as unethical for using placebo controls rather than providing a higher standard of care. The ensuing vigorous ethical debate helped shape current international guidelines on standards of care in clinical research4. Though the issue of standard of care remains controversial,5 current ethical guidelines outline when use of a placebo is justifiable. However, they provide limited help to a researcher having to choose among study designs with complex trade-offs among multiple criteria, including scientific validity, social value, protection of participants, and feasibility6. A study design that satisfies one criterion may fall short on another, and there is little guidance on how to balance these sometimes conflicting criteria.

These trade-offs are illustrated by considering how to design a trial to determine the efficacy of a licensed vaccine to prevent human papilloma virus (HPV) infection in HIV-infected adolescent girls living in lower income countries without current access to the vaccine. Vaccination against HPV is recommended by the World Health Organization (WHO)7 for all female adolescents (but not males), including those who are HIV-infected. Despite this recommendation, governments with limited health care dollars may be reluctant to add the costly (currently $130/dose8) vaccine to their national program. The choice is especially complicated for countries with high HIV prevalence, as efficacy of the HPV vaccine has not been studied in HIV-infected people and there are numerous examples of vaccines that are safe but less immunogenic or efficacious in HIV-infected individuals9. Adoption of a vaccine that might be less effective for a substantial proportion of recipients would waste money and divert resources that should be spent on more effective prevention strategies.

After providing some additional information on HPV and HPV vaccines, we describe alternative study designs which vary in how well they achieve balance in satisfying the criteria of scientific validity, social value, participant protection, and feasibility. We then show how existing ethical guidance (The Declaration of Helsinki10 and The Council of the International Organization of Medical Sciences (CIOMS11) is largely insensitive to these trade-offs. This case illustrates the need for improved ethical guidance that is better calibrated for research and public health decision-making.

BACKGROUND

HPV is the most common sexually transmitted disease in the world. Persistent HPV infection with certain high-risk strains is the most important risk factor for development of cervical and other anogenital dysplasias (pre-cancerous cells) and subsequent cancers in women12. High-risk HPVs and cervical cancer are common among women in lower income countries13. HIV-infected women are less able to clear HPV infection and more likely to develop cervical cancer than HIV-uninfected women14. Although there are some data on antibody levels in HIV-infected individuals, it is unclear whether a protective effect can be extrapolated from these data, and HPV vaccines do not have proven efficacy in HIV-infected individuals.

Gardasil, the Merck quadrivalent HPV vaccine (qHPV), administered as a three-dose series over six months15, is safe, immunogenic, and in HIV-uninfected women, prevents more than 98% of middle and high-grade cervical dysplasia caused by the HPV serotypes included in the vaccine16. Persistence of protection has been documented for at least five years17. Typically once a vaccine is proven efficacious in individuals without HIV infection, extension of its use to HIV-infected people depends only on demonstrating safety and immunogenicity. This, however, ignores the potential for wide variation in the extent or duration of protection in immune compromised individuals. A further complication is that for some vaccines, including qHPV, the level of immune response required to provide protection against disease (the correlate of protection) is not known. Without knowing this immune correlate, researchers cannot definitively infer that a vaccine will be efficacious in HIV-infected persons without recourse to a study using the efficacy endpoint of HPV infection18.

Safety and immunogenicity of qHPV have been established in HIV-infected girls and boys 7–12 years of age in the United States (US) who were either receiving highly active antiretroviral therapy (HAART) or were immunologically robust (CD4% ≥25%) and did not require HAART19. Importantly, immunogenicity was 30% to 50% lower than in HIV-uninfected girls and boys for two of the four serotypes, raising the possibility that the vaccine might be less effective or offer protection for a shorter time in HIV-infected vaccinees and establishing a need for efficacy data.

POSSIBLE STUDY DESIGNS

Given the importance of establishing the efficacy of qHPV in HIV-infected girls for public health decision-making, we consider alternative study designs. We assume the studies would be conducted only in countries with high HIV prevalence, where the vaccine is not available, and where local Institutional Review Boards deem the study to be acceptable and of sufficient value to the country. Studies would counsel participants in safe sex and offer condoms. It should be noted that even if the vaccine was shown to be insufficiently efficacious to justify adoption as standard of care, a properly designed study might provide enough information to determine the correlate of protection for qHPV. This in turn would facilitate design of future studies addressing new vaccine formulations and strategies for lower income settings. Negative evidence would also allow governments to reallocate resources planned for the HPV vaccine and could lead to additional research to enhance efficacy in this population.

All proposed study designs discussed below would follow participants for three years with a study endpoint of a new persistent HPV infection with any of the four serotypes present in qHPV which was not present at study entry, an endpoint which has been used in multiple previous studies20 (the gold standard of using PAP smears is not feasible in resource-poor settings). The age range reflects the Advisory Committee on Immunization Practices (ACIP)21 recommendation to vaccinate in early adolescence before sexual debut. Both perinatally and horizontally HIV-infected girls would be enrolled just before or soon after they become sexually active.

It is likely that there would be sufficient numbers of HPV-negative individuals in the target population. Very few perinatally HIV-infected girls would be HPV-infected before sexual initiation. Horizontally HIV-infected girls would likely have become HIV-infected through sex, but it is unlikely they would be HPV-positive at enrollment for all four HPV types in the qHPV vaccine.

Each design is discussed in terms of the following criteria: scientific validity, social value, protection of participants, feasibility (represented by sample size) and compliance with existing ethical guidance (summarized in Table 1). It is important to note that only the design described in Option 1 would allow direct measurement of vaccine efficacy22 in the target population.

Table 1

Summary of ethical principles for alternative study designs for assessing efficacy of an HPV vaccine in HIV-infected adolescent girls in resource-limited settings

DesignScientific validityLocal social value of the dataProtection of participantsFeasibility
Compliance with ethical guidance under Helsinki and CIOMS
Estimated HPV infection rateN/arm (evaluable)4
1. Superiority: 3 doses vs placebo+crossoverCleanest dataHighestParticipants in active arm protected from start, placebo arm will receive benefit of vaccine in 3 years3 dose ratePlacebo rateNot permissible under Helsinki; unclear whether allowed under CIOMS
2%10% (80% VE3)162 1
2%20% (90% VE)57
4%14% (70% VE)147
4%40% (90% VE)25
2a. Open label of 3 doses vs historical controlsDepends on interpretable comparison of efficacy to historical controlsLimitedAll participants equally protected3 dose rateWidth 95% CIPermissible under both
2%± 3%200
4%± 4%200
2b. Superiority of 4 vs 3 doses (historical controls)Depends on interpretable comparison of efficacy to historical controlsLimited—4 doses even less affordableBoth arms receive same or greater protection as developed countries4 dose rate3 dose ratePermissible under both
2%4%1239 1
2%6%524
4%6%1962
4%8%602
2c. Non-inferiority: 3 vs 2 doses (historical controls)Depends on interpretable comparison of efficacy in 3 dose arm to historical controlsArguably highest, depending on whether 2 or 3 doses plausibly offer protectionBoth arms receive some protection3 dose rate2 dose rateUnclear whether permissible under Helsinki, clearly permissible under CIOMS
2%4%8622
2%6%255
4%6%1547
4%8%411
3. Adding HIV uninfected controls to 2a–2cBetter comparison armMore relevant controls which improves decision- makingAll participants equally protectedPermissible under both for 2a and 2b, unclear whether permissible under Helsinki for 2c
1Difference in event rates between standard 3-dose regimen and comparator arm interpreted as ‘clinically meaningful difference’ between arms that study designed to detect. Sample size assumes a 2-sided Type I error of 0.05
2Difference in event rates interpreted as ‘acceptable decrease’ in efficacy of 2-dose regimen relative to 3-dose regimen. Sample size assumes a 1-sided Type I error 0.025.
3Vaccine efficacy (VE)
4Needs to be inflated to account for loss-to-follow-up, interim analyses, exposure to HPV by age etc

Option 1 - Superiority, Placebo Controlled, Crossover Design

The first option is to randomize HIV-infected adolescent girls to either three doses of qHPV or placebo. At the end of follow-up, participants would be unblinded and the placebo group would receive the vaccine, so that ultimately all participants would receive qHPV.

Scientific validity

Randomization would help ensure that placebo and vaccine groups were similar not only with respect to HIV clinical status, but also for sexual behavior and exposure to HPV during the study, allowing unbiased estimates of vaccine efficacy. Because the HIV-infected girls in the study would likely have other co-morbidities, having a placebo group would also be the best way to determine whether adverse events might be attributable to the vaccine.

Feasibility

Because of the relatively large anticipated vaccine efficacy, sample sizes are not overly large relative to competing study designs (see Table 1).

Social value

This design will provide definitive in-country safety, immunogenicity and efficacy information, giving governments in areas of high HIV prevalence the most easily interpretable data for making a decision about adopting qHPV vaccination.

Protection of participants

The major concern with this design is use of a placebo when the efficacy of qHPV has been established in immune competent populations. Each participant would have a 50/50 chance of receiving qHPV immediately and those in the placebo group would receive the vaccine after three years of follow-up. The ethical cost of this approach is that during the initial three years of the study, placebo recipients would not have whatever protection against HPV infection would be afforded by qHPV, which could be considered a harm. It is widely accepted in the medical ethics literature that researchers have a duty to provide the standard of care, meaning that it is not sufficient to simply ensure that participants are made better off than they would be otherwise with study participation23. One reason for this is that all of us, including researchers, have a duty to perform easy rescues. The classic example is an adult walking by a child drowning in a shallow pool of water, who could easily be pulled to safety. Just as the adult has the duty to pull the child out of the water even if there is a slight cost to him/her, it seems clear that we all have duties to help the people we interact with when we can help them greatly at minimal cost to ourselves24.

Applying this duty to the present case, we need to determine whether participants are likely to benefit significantly from receiving vaccine instead of placebo and how much it would cost researchers to do so. The cost here is not merely the monetary cost of providing the vaccine, but how critical it is to use a placebo control in order to yield scientifically valid results. As discussed in the Option 2 designs, a priori we know that using no control or an active control may yield biased and less interpretable results, in which case study participants could be burdened by trial participation for no good reason, and the opportunity to produce valuable data that could have saved others will have been lost.

The question then becomes - what is the minimum standard of care/prevention that investigators have a duty to provide to participants? Wendler and colleagues say that based on the duty to perform easy rescues, the answer depends on the cost of providing a higher standard and the relevance of the study question to the local population. This suggests that if the qHPV study was very relevant for the host community and the chance of yielding valid results is significantly reduced without a placebo control, then the study may be acceptable. On the other hand, some claim that investigators have to offer the best proven interventions to participants. This would mean providing qHPV to all participants, but only if certainty about the efficacy of the vaccine in HIV-infected adolescents was sufficiently high. Still others argue there is some threshold between the best proven interventions and what is locally available, though what this middle ground is in specific cases is hard to determine25. There is no universal agreement on the answers to these questions.

Existing ethical guidance

Several guidelines address the ethics of using a placebo in a clinical trial. The two most prominent are the World Medical Association’s Declaration of Helsinki and the guidelines issued by CIOMS. The Declaration of Helsinki permits placebo controls when (1) there is no current proven intervention or (2) the use of a placebo is scientifically necessary and the placebo group will not be exposed to risk of serious or irreversible harm26. CIOMS guidance has a similar general rule about use of a placebo, but departs from Helsinki by providing an exception that placebo may be appropriate: ‘… in a country1 or community in which an established effective intervention is not available and unlikely in the foreseeable future to become available, usually for economic or logistic reasons…’ and where the intervention is responsive to local health needs27.

The use of placebo appears to violate both conditions outlined in the Declaration of Helsinki. As mentioned earlier, although there is no current intervention that has been proven to prevent HPV in HIV-infected individuals, the question is whether we can extrapolate from the existing data in individuals who do not have HIV and from the safety and immunogenicity data in HIV-infected individuals. The WHO recommendations imply that the evidence is sufficient. There also appears to be a risk of serious and irreversible harm from the use of placebo. Any participants in the placebo arm who do not become infected with HPV would be given the vaccine at the end of the study. During the first three years of the HPV study, however, placebo recipients could be infected with HPV, increasing their likelihood of developing cervical dysplasia and even cancer.

The Declaration of Helsinki guidance, however, has been criticized as overly demanding and out of step with other guidance documents, including CIOMS28. Under CIOMS, use of a placebo is not permitted if an intervention is not available in a country and unlikely to be made available in the near future, since the study population would be unlikely to benefit from the data they help generate, leading to concern that the host community is being exploited29. It is very difficult to predict whether the HPV vaccine would become available in the near future in the host countries. Given the high cost of the vaccine, there is reason to believe it may not become widely available. However, the pharmaceutical company has raised the possibility of tiered pricing, although this is not yet in effect, and actual pricing levels would affect the ability of countries to adopt use of the vaccine.

These international ethical guidelines are influential but take different approaches and are subject to important criticisms. They raise questions without easy answers. What counts as a proven intervention? If an intervention like the HPV vaccine is proven in the general population, but there is uncertainty about extrapolating the results to a subpopulation of HIV-infected adolescent girls, can further research be accomplished under the restrictions of the existing ethical guidance? What counts as a risk of serious or irreversible harm? Is the fact that some girls who receive the placebo will likely become infected with HPV sufficiently serious? Finally, how can researchers and ethics committees predict what intervention will become available ‘in the foreseeable future’? Given the high cost of the HPV vaccine, it is difficult to imagine it becoming widely available in the near future. Nevertheless, many high-cost interventions for HIV-infected patients in lower income settings were introduced because of robust placebo-controlled trials that fueled advocacy which ultimately led to their availability, e.g. combination antiretrovirals30.

Option 2 – Designs Without Untreated Controls

Since the vaccine is likely to be at least somewhat efficacious in HIV-infected adolescents, the designs proposed in this section are less problematic in terms of complying with existing guidelines and protecting participants, as they provide vaccine to all participants. All these designs, however, may compromise scientific validity to some degree in terms of interpretability of study results, which in turn reduces social value. Equivocal results may lead a policy-making body to not adopt an efficacious vaccine or to adopt a non-efficacious vaccine and thereby use valuable resources that might be better spent on other public health interventions. As previously discussed, this raises the concern that a study that does not violate ethical principles regarding protection of participants but is unlikely to produce readily interpretable results, may still be unethical.

The primary difficulty with scientific validity for designs under Options 2 and 3 is identifying adequate control groups. Relying on comparing the rates of HPV infection collected in the HIV-infected subjects in the study with rates in historical HIV-uninfected controls collected at different times or in different places is problematic. If infection rates are comparable to or lower than those observed in historical controls, this might mean the vaccine was effective in the study population, or that rates of sexual activity and HPV exposure in the study population were lower than in the historical controls, or there were other unmeasured cofactors influencing persistent infection that were different in the study and control populations. If higher rates of infection are observed in the study group this could also be due to differences in exposure or other unknown factors. In addition, rates of infection with no vaccine may differ substantially between HIV-infected and HIV-uninfected adolescent females. Significantly, these designs allow no direct measurement of vaccine efficacy in the HIV-infected group.

Option 2A – Open Label Uncontrolled Trial Administering Three Vaccine Doses to HIV-Infected Girls

This simple design would enroll a cohort of HIV-infected adolescent girls, administer three doses of qHPV, collect information on safety and immunogenicity, and follow them long enough to estimate the proportion developing persistent HPV infection. This proportion would be compared to published estimates from trials in older women with the same endpoint.

Scientific validity

Although information on safety, immunogenicity, and endpoint rates would be collected, the difficulty would be in comparing each outcome to data collected in either HIV-infected (only safety and immunogenicity) or uninfected (all outcomes) historical controls. The immunogenicity information obtained would be useful when placed in the context of reference comparison groups but is likely to be of limited value in the absence of correlates of protection. Furthermore, the comparison of endpoint rates assumes similar exposure to HPV and matching for other factors between groups, which is unlikely to be achievable outside of a randomized trial.

Feasibility

The study would require a relatively small sample size.

Social value

The uncertainty of scientifically interpretable results would result in governments having less optimal information for assessing the value of the vaccine in their setting, increasing the likelihood of an incorrect decision about adopting the vaccine as standard of care.

Protection of participants

This design does not raise significant concerns about participant protection.

Current ethical guidance

This design does not appear to violate existing ethical guidance.

Option 2B – Superiority Design Randomizing HIV-Infected Girls to Four vs. Three Doses of Vaccine

The licensed and recommended dosing of qHPV is three doses. As shown in an earlier trial31, immunogenic responses to the vaccine in HIV-infected girls and boys were lower than those in HIV-uninfected children. A follow-up study of these subjects in the US suggested that four doses yield higher immunologic responses32. On this basis, a potential design would randomize HIV-infected girls to receive either three or four doses of vaccine, to determine if infection rates in the four dose regimen would be lower.

Scientific validity

Immunogenicity for both dose regimens in the lower income setting could be compared to immunogenicity in the ongoing study in the US and relative efficacy of the two dose regimens within the study could be measured. Regardless of rates of HPV infection in the two arms, assessing vaccine efficacy still requires comparison with historical controls, with the associated difficulties in interpretation described previously.

Feasibility

The sample size required to detect differences in efficacy between three and four doses would probably be unfeasibly large (see Table 1), since rates of HPV infection would likely be low and differences small.

Social value

Given the higher cost of giving people four doses, this design would be less relevant to the needs of host countries facing significant resource constraints.

Protection of participants

Additional monitoring for safety of a fourth vaccine dose would be needed, but this design does not raise significant concerns about participant protection.

Current ethical guidance

This design does not appear to violate existing ethical guidance.

Option 2C – Non-Inferiority Design Randomizing HIV-Infected Girls to Two vs. Three Doses of Vaccine

This study design would randomize HIV-infected girls to either two or three doses of qHVP, based on observational data from recent large trials showing that efficacy after one or two doses of a bivalent HPV vaccine over a median follow-up of 4.2 years was similar to that of the full three-dose series33, and immunogenicity responses were similar to those in HIV-uninfected participants receiving qHPV34. Showing that two doses are as effective as (or minimally less effective than) the recommended schedule could result in large savings. The two-dose regimen would be considered acceptable if the proportion of participants developing persistent HPV infection was within the non-inferiority margin of an ‘acceptable’ difference from the three-dose regimen and if the rate of infection with the three-dose regimen was deemed satisfactory relative to historical (or possibly concurrently-enrolled HIV-uninfected girls – see next section) controls.

Scientific validity

Assessment of the adequacy of the three-dose regimen using control data is vitally important to interpretation of the study results and subject to the difficulties in interpretation outlined previously.

Feasibility

Sample sizes are smaller than for the superiority design described in Option 2b (Table 1) but are sensitive to the choice of non-inferiority margin (i.e, how much less protection with two doses is acceptable). It might also be necessary to extend the length of study follow-up since short term protection might be similar in the two arms but longer-term protection might wane more quickly in the two dose regimen.

Protection of participants

Administering only two doses to half the participants could be deemed ethically problematic since it is not the recommended schedule and it could be argued that participants are being randomized to less than the best proven standard of care. Additionally, HIV-infected youth in the US study had lower humoral responses to qHPV than did HIV-uninfected youth, and it has been established for many other vaccines that HIV-infected patients need a greater number or strength of doses to achieve an optimal response35. These concerns could be alleviated by existing data referred to above (albeit in an HIV-uninfected population) that suggest favorable performance of a two dose regimen and also by vaccinating those in the two-dose group at the end of follow-up, acknowledging the longer risk period of HPV acquisition (pre third dose) and that the interval between the second and third vaccinations would be longer than recommended with possible effects (beneficial or not) on immunogenicity36. These factors raise concern about the appropriateness of selecting an HIV-infected population for an efficacy trial of a reduced-dose regimen of HPV vaccine.

Social value

Adopting a two-dose regimen might involve a trade-off between the potential morbidities from decreased protection (a function of the non-inferiority margin chosen for trial design) and the reduction in costs incurred by administering only two vaccinations. Governments might need cost-benefit analyses to make a final decision, but if acceptable, results from this study could result in significant cost savings and would be very relevant for policy makers distributing limited health care resources.

Current ethical guidance

This design would be permissible under CIOMS. It is unclear whether the design is permissible under Helsinki because the two-dose regimen which protected women without HIV infection might offer less protection to HIV-infected girls.

Option 3 - Designs Without Untreated Controls Including Cohorts of HIV-Infected and HIV-Uninfected Girls

Study designs described in Option 2 could be expanded to include cohorts of HIV-uninfected girls.

Scientific validity

This would be increased by studying individuals in all study arms from a similar background with the same study endpoint. However, even when the HIV-uninfected subjects are drawn from the same population, they may differ from the HIV-infected participants at study entry, and their sexual behavior and exposure to HPV may differ during follow-up, biasing comparisons of HPV infection rates. Despite this limitation, they are the best reference group for comparison of incident HPV infection.

Feasibility

Including HIV-uninfected groups increases sample size, possibly to the extent that the trial will be too expensive to implement.

Protection of participants

No significant concerns are raised by including HIV-uninfected girls under designs 2a and 2b. Under a design like that presented in 2c, where half the participants would receive two rather than three doses, there would be reservations about providing HIV-infected participants with inadequate protection as discussed above, but these would be balanced by evidence of comparable efficacy for the two-dose series in HIV-uninfected girls from observational studies.

Social value

With the increased scientific value of enrolling controls drawn from the same population, governments would have better (but not ideal) information on which to base their decision about usefulness of the vaccine.

Current ethical guidance

It is unclear whether randomizing HIV-infected women to receive two doses of the qHPV vaccine would be permissible under the Declaration of Helsinki because the two-dose regimen which protected women without HIV infection might offer less protection to HIV-infected girls.

DISCUSSION

The current state of the debate over standard of care may not be helpful in illuminating situations increasingly faced by decision-makers, sponsors and clinical researchers. The qHPV vaccine example illustrates that existing ethical guidance not only fails to help them navigate amongst different trial designs that favor different ethical criteria, but in some cases, hinders research that would provide local governments with the clearest information on which to base public health decisions.

Our analysis suggests that to assess the efficacy of an HPV vaccine in HIV-infected girls, Option 2c—randomizing HIV-infected girls to two versus three doses of qHPV—does the best across the various criteria. This design has the best scientific validity among the designs with no untreated controls, clear social value for the host countries, is feasible, and provides some degree of protection to all participants. Although it is not clear whether this design would satisfy the Declaration of Helsinki criteria, Helsinki has been widely criticized for treating participant protection as trumping all other considerations, and failing to recognize other ethical criteria sufficiently. It is possible, however, that Option 2c will be unable to produce scientifically valid results. The least problematic design in terms of scientific validity is Option 1 - the placebo-controlled crossover trial randomizing HIV-infected girls to three doses of the vaccine or placebo, but this is also the most problematic approach on the criterion of participant protection.

Stepping back from existing ethical guidance and criteria, there are other ways to evaluate this research. One way would be to take a utilitarian approach and assume that whatever study design serves the greater good should be adopted37. However, almost everyone would agree that there must be some limits to utilitarianism in research. Some of the research scandals of the past may have been justifiable on utilitarian grounds, but are almost universally seen as unethical today. On the other end of the spectrum, guidance documents like the Declaration of Helsinki value participant protection above all else. Yet this approach simply does not fit with the primary goal behind the conduct of research—the goal of developing generalizable knowledge for future patients.

Neither utilitarianism nor relying on absolute participant protection is sufficiently nuanced to capture the complex ethical trade-offs involved in the conduct of research, and existing ethical guidance has significant limitations. We argue that, to be useful to those involved in the conduct of research, ethical guidance should recommend that determining how to design a research study amongst competing options, researchers and sponsors should choose the design that achieves the best balance regarding scientific validity, social value, feasibility, and protection of participants.

Acknowledgments

We would like to thank Alan Wertheimer, Doug MacKay, Nicola Barsdorf, and Rohan Hazra for their helpful comments and suggestions. Overall support for the International Maternal Pediatric Adolescent AIDS Clinical Trials Group (IMPAACT) was provided by the National Institute of Allergy and Infectious Diseases (NIAID) [U01 AI068632], the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), the National Institute of Mental Health (NIMH) [AI068632], and the Intramural Research Program of the NIH in the Warren G. Magnussen Clinical Center. This work was supported by the Statistical and Data Analysis Center at Harvard School of Public Health, under the National Institute of Allergy and Infectious Diseases cooperative agreement #5 U01 AI41110 with the Pediatric AIDS Clinical Trials Group (PACTG) and #1 U01 AI068616 with the IMPAACT Group. This project has also been supported with Federal funds from the National Institute of Allergies and Infectious Diseases, National Institutes of Health, Department of Health and Human Services, under Contract No. HHSN272200800014C. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the U.S. Government nor does mention of trade names, commercial practices, or organizations imply endorsement by the U.S. Government. Some of the authors are U.S. government employees who must comply with the NIH Public Access Policy, and the authors or NIH will deposit, or have deposited, in NIH’s PubMed Central archive, an electronic version of the final, peer-reviewed manuscript upon acceptance for publication, to be made publicly available no later than 12 months after the official date of publication.

Biographies

• 

Jane Lindsey is a Senior Research Scientist at the Center for Biostatistics in AIDS Research at the Harvard School of Public Health. She works as a Biostatistician at the Statistical and Data Analysis Centre for the International Maternal Pediatric Adolescent AIDS Clinical Trials group on clinical trials and observational studies in children with HIV.

• 

Seema Shah is a faculty member in the Clinical Center Department of Bioethics and has a joint appointment at the Division of AIDS. Her research focuses on the ethics of international research, the ethics of research with children, and the intersection of law and bioethics.

• 

Myron Levin is Professor of Pediatrics and Medicine (Director of Vaccine Clinical Trials) at the University of Colorado Anschutz Medical Campus. He is a clinical virologist who cares for HIV-infected children and has chaired four protocols within the International Maternal Pediatric Adolescent Clinical Trials network that investigated vaccine responses in these children.

• 

George Siberry is a Medical Officer at the Maternal and Pediatric Infectious Disease (MPID) Branch of the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (NIH) in Bethesda, Maryland (USA). He develops research studies focused on prevention, treatment and complications of pediatric and adolescent HIV infection through several domestic and international HIV research networks.

• 

Patrick Jean-Philippe is a medical officer at the Maternal Adolescent Pediatric Research Branch of the Division of AIDS (DAIDS), National Institute of Allergy and Infectious Diseases, National Institute of Health and a contractor for the Henry M. Jackson Foundation. His duties include safety and scientific oversight as well as protection of human subject participants in trials supported by the DAIDS.

References

1. Connor EM, Sperling RS, Gelber R, et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. NEJM. 1994;331:1173–1180. [Abstract] [Google Scholar]
2. Lurie P, Wolfe SM. Unethical trials of interventions to reduce perinatal transmission of the human immunodeficiency virus in developing countries. NEJM. 1997;337:853–856. [Abstract] [Google Scholar]
3. Hawkins JS, Emanuel EJ, editors. Exploitation and developing countries: The ethics of clinical research. Princeton, NJ: Princeton University Press; 2008. [Google Scholar]
4. Macklin R. Double standards in medical research in developing countries. Cambridge: Cambridge University Press; 2004. [Google Scholar]
5. Philpott S, Schuklenk U. A Study that Should Not Have Been Done. [Accessed 14 Feb 2013].];Bioethics Forum, Hastings Center Report. 2010 41:3. Available at: http://www.thehastingscenter.org/Bioethicsforum/Post.aspx?id=4626&blogid=140. [Google Scholar]
6. Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical? JAMA. 2000;283:2701–11. [Abstract] [Google Scholar]
7. WHO recommends ‘that routine HPV vaccination should be included in national immunization programmes, provided that: prevention of cervical cancer or other HPV-related diseases, or both, constitutes a public health priority; vaccine introduction is programmatically feasible; sustainable financing can be secured; and the cost effectiveness of vaccination strategies in the country or region is considered’. Specifically for HIV-infected females, WHO states: ‘…the potential benefit of vaccination in this group is particularly great owing to their increased risk of HPV-related disease, including cervical cancer…. Concerns about safety or reduced efficacy among females who may be infected with HIV should not defer the initiation of large-scale HPV immunization.’ World Health Organization. Human papillomavirus vaccines. WHO position paper. Weekly Epidemiol Rec. 2009;84:118–131. [Abstract] [Google Scholar]
8. CDC. [Accessed 14 Feb 2013];HPV Vaccine Information for Young Women - Fact Sheet. 2012 Jul 17; Available at: http://www.cdc.gov/std/hpv/stdfact-hpv-vaccine-young-women.htm.
9. Abzug MJ. Vaccination in the immunocompromised child: a probe of immune reconstitution. PIDJ. 2009;28:233–236. [Abstract] [Google Scholar]Siberry GK, Warshaw MG, Williams PL, et al. Safety and immunogenicity of quadrivalent meningococcal conjugate vaccine in 2- to 10-year-old human immunodeficiency virus-infected children. PIDJ. 2012;31:47–52. [Europe PMC free article] [Abstract] [Google Scholar]Madhi SA, Adrian P, Kuwanda L, et al. Long-term immunogenicity and efficacy of a 9-valent conjugate pneumococcal vaccine in human immunoedeficient virus infected and non-infected children in the absence of a booster dose of vaccine. Vaccine. 2007;25:2451–2457. [Abstract] [Google Scholar]Madhi SA, Kuwanda L, Saarinen L, et al. Immunogenicity and effectiveness of Haemophilus influenzae type b conjugate vaccine in HIV infected and uninfected African children. Vaccine. 2005;23:5517–5525. [Abstract] [Google Scholar]Madhi SA, Petersen K, Khoosal M, et al. Reduced effectiveness of Haemophilus influenzae type b conjugate vaccine in children with a high prevalence of human immunodeficiency virus type 1 infection. PIDJ. 2002;21:315–321. [Abstract] [Google Scholar]Landrum ML, Hullsiek KH, Ganesan A, et al. Hepatitis B vaccination and risk of hepatitis B infection in HIV-infected individuals. AIDS. 2010;24:545–555. [Europe PMC free article] [Abstract] [Google Scholar]
10. World Medical Assocation. [Accessed 14 Feb 2013];Declaration of Helsinki-Ethical Principles for Medical Research Involving Human Subjects. 2008 Oct 28; Available at: http://www.wma.net/en/30publications/10policies/b3/index.html.pdf?print-media-type&footer-right=[page]/[toPage]
11. CIOMS. International ethical guidelines for biomedical research involving human subjects. Geneva, Switzerland: CIOMS; 2002. [Abstract] [Google Scholar]
12. Wheeler CM. Natural history of human papillomavirus infections, cytologic and histologic abnormalities, and cancer. Obstet Gynecol Clin North Am. 2008;35:519–536. [Abstract] [Google Scholar]
13. Ferlay J, Shin HR, Bray F, et al. GLOBOCAN 2008, Cancer Incidence and Mortality Worldwide: IARC CancerBase No 10. Lyon, France: International Agency for Research on Cancer; 2008. [Google Scholar]
14. Moscicki AB, Ellenberg JH, Crowley-Nowick P, et al. Risk of high-grade squamous intraepithelial lesion in HIV-infected adolescents. JID. 2004;190:1413–21. [Abstract] [Google Scholar]
15. Markowitz LE, Dunne EF, Saraiya M, et al. Quadrivalent Human Papillomavirus vaccine: Recommendations of the Advisory Committee on Immunization Practices (ACIP) MMWR. 2007 Mar 23;56(RR-2):1–24. [Abstract] [Google Scholar]
16. The Future II Study Group. Quadrivalent vaccine against human papillomavirus to prevent high-grade cervical lesions. NEJM. 2007;356:1915–1927. [Abstract] [Google Scholar]
17. Romanowski B. Long term protection against cervical infection with the human papillomavirus: review of currently available vaccines. Hum Vaccin. 2011;7:161–169. [Abstract] [Google Scholar]
18. Plotkin SA. Correlates of vaccine-induced immunity. CID. 2008;47:401–409. [Abstract] [Google Scholar]
19. Levin MJ, Moscicki AB, Song LY, et al. Safety and immunogenicity of a quadrivalent human papillomavirus (types 6, 11, 16, and 18) vaccine in HIV-infected children 7 to 12 years old. JAIDS. 2010;55:197–204. [Europe PMC free article] [Abstract] [Google Scholar]
20. Lu B, Kumar A, Castellsague X, et al. Efficacy and safety of prophylactic vaccines against cervical HPV infection and diseases among women: a systematic review and meta-analysis. BMC Infect Dis. 2011 Jan;12:11–13. [Europe PMC free article] [Abstract] [Google Scholar]Munoz N, Manalastas R, Jr, Pitisuttithum P, et al. Safety, immunogenicity, and efficacy of quadrivalent human papillomavirus (types 6, 11, 16, 18) recombinant vaccine in women aged 24–45 years: a randomised double-blind trial. Lancet. 2009;373:1949–1957. [Abstract] [Google Scholar]Chen HC, Schiffman M, Lin CY, et al. Persistence of type-specific human papillomavirus infection and increased long-term risk of cervical cancer. JNCI. 2011;103:1387–1396. [Europe PMC free article] [Abstract] [Google Scholar]
21. CDC. Quadrivalent Human Papillomavirus Vaccine: Recommendations of the Advisory Committee on Immunization Practices (ACIP), 2007. MMWR. 2007;56(RR-2) [Abstract] [Google Scholar]
22. Vaccine efficacy is defined as:
VE=Rp-RvRp∗100%
where Rp is the proportion or incidence rate of participants developing persistent HPV infection with no vaccination and Rv is the corresponding rate in the group receiving the vaccination.
23. Lurie & Wolfe, op. cit. note 2, pp. 853–856; Wendler D, Emanuel EJ, Lie RK. The standard of care debate: can research in developing countries be both ethical and responsive to those countries’ health needs? AJPH. 2004;94:923–928. [Abstract] [Google Scholar]
24. Hawkins & Emanuel, eds., op. cit. note 3.
25. Lie RK, Emanuel E, Grady C, Wendler D. The standard of care debate: the Declaration of Helsinki versus the international consensus opinion. J Med Ethics. 2004;30:190–193. [Europe PMC free article] [Abstract] [Google Scholar]Hyder AA, Dawson L. Defining standard of care in the developing world: the intersection of international research ethics and health systems analysis. Developing World Bioethics. 2005;5:142–152. [Abstract] [Google Scholar]London AJ. Equipoise and international human-subjects research. Bioethics. 2001;15:312–332. [Abstract] [Google Scholar]
26. Declaration of Helsinki-Ethical Principles for Medical Research Involving Human Subjects, op. cit. note 10.
27. CIOMS, note 11.
28. Lie, Emanuel, Grady et al, op. cit. note 25, pp. 923–928.
29. CIOMS, op. cit. note 11.
30. Cleaton-Jones PE. An ethical dilemma: availability of antiretroviral therapy after clinical trials with HIV-infected patients are ended. BMJ. 1997;314:887–888. [Europe PMC free article] [Abstract] [Google Scholar]
31. Levin, Moscicki, Song et al, op. cit. note 19, pp. 197–204.
32. Weinberg A, Song LY, Fenton T, et al. Response to protective vaccines. Conference on Retroviruses and Opportunistic Infections; Seattle, WA. 2012. [Google Scholar]
33. Kreimer AR, Rodriguez AC, Hildesheim A, et al. Eurogin. Lisbon; Portugal: 2011. Evaluation of the efficacy of fewer than three doses of a bivalent HPV 16/18 vaccine. [Europe PMC free article] [Abstract] [Google Scholar]
34. Neuzil KM, do Canh G, Thiem VD, et al. Immunogenicity and reactogenicity of alternative schedules of HPV vaccine in Vietnam: a cluster randomized noninferiority trial. JAMA. 2011;305:1424–1431. [Abstract] [Google Scholar]
35. Madhi, Petersen, Khoosal et al, op. cit. note 9, pp. 315–231; Landrum, Hullsiek, Ganesan et al, op. cit. note 9, pp. 545–555 Advisory Committee on Immunization Practices. Updated recommendations for use of meningococcal conjugate vaccines. MMWR. 2011;60:72–76. [Abstract] [Google Scholar]
36. Neuzil, Canh, Thiem et al, op. cit. note 34, pp. 1424–1431.
37. Goodin RE. Utilitarianism as a public philosophy. Cambridge: Cambridge University Press; 1995. [Google Scholar]

Full text links 

Read article at publisher's site: https://doi.org/10.1111/dewb.12028

Read article for free, from open access legal sources, via Unpaywall: https://europepmc.org/articles/pmc3716837?pdf=renderUnpaywall

Citations & impact 

Impact metrics

2
Citations
Jump to Citations

Article citations

Similar Articles 

To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.

Funding 

Funders who supported this work.

NIAID NIH HHS (9)

PHS HHS (1)