|
Development Of Priority Vaccines For Disease-Endemic Countries: Risk And Benefit
Decisions regarding vaccine regulation and use made by institutions in industrialized countries can have an unintended impact on vaccines’ availability in disease-endemic countries. However, regulatory and programmatic decision making by such countries, taking into consideration local risks and benefits, requires adequate resources, both human and financial. Such differing risk-benefit determinations between industrialized and disease-endemic countries will increase product divergence. We propose a single universal standard for risk assessment, based on maximizing net benefit, and an action plan to improve access to priority vaccines through a more robust determination of risk and benefit.
Many issues affect access to priority vaccines for the developing world. One issue in particular has not received sufficient attention: the interpretation of risk versus benefit often made, sometimes inadvertently and often inappropriately, for the developing world by institutions in industrialized countries. Often this occurs by default because of a lack of sufficiently strong frameworks in disease-endemic countries to make complex regulatory and programmatic decisions. Another factor complicating the risk-benefit impact of vaccines is product divergence. Vaccine products may diverge between industrialized and disease-endemic countries because of differences in disease burden, health characteristics of the population, vaccine efficacy and safety, or economic power. Assessment of risk and benefit based on one set of assumptions regarding these factors may not be applicable under a different set of assumptions. Ethical considerations in the determination of risk and benefit may also differ greatly. Applying standards relevant to the use of vaccines in industrialized countries may not be appropriate for some target populations. Clear criteria must be established to support such determinations. This paper provides case studies highlighting the importance of risk-benefit considerations in the context of local vaccine use and proposes approaches to address the issues raised in these studies.
Regulatory and programmatic decisions affect vaccines’ availability. Regulatory decisions relate to inputs on testing and approval of products by a national regulatory authority (NRA) as well as, in some cases, institutional review boards (IRBs) or ethics committees. Programmatic decisions involve policy determinations on vaccine use. Many disease-endemic countries lack capacity in both areas, which means that they must depend on decisions made in industrialized countries. When this happens, the determination of risk and benefit may not be appropriate for the local situation. Regulatory decisions. When the product is already in use The NRA should be involved early in the vaccine development process—for example, in review of preclinical data, authorization of clinical trials, and inspection of the production facility. This involvement continues with regulatory approval of the product, including review of clinical trial data and postmarketing activities such as lot release, ongoing assurance of compliance with current Good Manufacturing Practices (cGMP), monitoring of field experience of products, and approval of manufacturing changes. For the United States, the Food and Drug Administration (FDA) authorizes all clinical trials through an investigational new drug (IND) process and continues its involvement throughout a product’s life cycle.1 Historically, new vaccine products have come to the developing world following a thorough regulatory review and often a ten- to twenty-year history of use.2 Many NRAs in developing countries explicitly include the provision that a product must be approved and used in the country of production to be eligible for regulatory approval in their country.3 Often this decision is based on the assumption that the vaccine product will be safe and effective in the disease-endemic country as well, even though in most cases, the original NRA gave no such consideration to the vaccine product. Yet clinical trial data must be specifically reviewed to ensure that conditions that could interfere with safety and efficacy, such as different immunization schedules or disease burdens, have been considered. In the prequalification of vaccines for purchase by United Nations (UN) agencies for global immunization programs, the World Health Organization (WHO) performs this task on behalf of the developing world through expert review of the clinical trial data and requests additional data if necessary.4 When the product is developed to prevent prevalent developing-world diseases. For vaccines to be developed against diseases prevalent in the developing world, the route outlined above might not apply. The product might never be approved for use in the country of production because of projected lack of demand. Thus, pivotal clinical trials to demonstrate safety and efficacy could occur in disease-endemic countries, often with limited or no testing elsewhere. Such countries may have to take responsibility for approving and evaluating data from these trials with little or no historical guidance on the salient issues. Vaccine developers might choose to submit a file for such a product through the FDA’s IND process, even though the endpoints and indications might not be relevant to the U.S. situation because of a lack of disease burden. Other factors that could adversely affect the time frames, costs, or trial outcomes of such products include the following. (1) Experience with BCG: Bacille Calmette-Guérin (BCG) vectors modified for use in new tuberculosis vaccine constructs could present a regulatory challenge for an NRA in a country with relatively low tuberculosis disease burden and little experience with BCG. Decisions on the clinical development pathway might thus be cautious. For example, although it might appear justified on grounds of absolute safety, it is difficult to conduct a trial first in BCG-naïve populations before proceeding to the general population. Such potential trial participants might be rare in a disease-endemic country, thus severely limiting the population eligible to participate and possibly resulting in delays to approval of a needed product. (2) Treatment ethics for HIV vaccine trial subjects: The initiation of preventive HIV vaccine trials in high-incidence countries has been delayed by debate over the current wording of Paragraph 30 of the Declaration of Helsinki, which states, "At the conclusion of the study, every patient entered into the study should be assured of access to the best proven prophylactic, diagnostic, and therapeutic methods identified by the study." Under pressure from the U.S. government and the pharmaceutical industry to change the wording, the World Medical Association has proposed to postpone a decision. Meanwhile, trials have been delayed pending this decision. The U.S. government is not of one mind, however. An FDA official stated that Paragraph 30 needs either clarification or modification, while the HIV Vaccine Trials Network (VTN), supported by the U.S. National Institutes of Health (NIH), has resolved to provide antiretroviral treatment for all participants in HIV vaccine trials that become infected over the course of the trial.5 (3) Use of placebo controls with Haemophilus influenzae type b (Hib) and pneumococcal conjugate vaccines (PCV7): In the Gambia (West Africa), the Gambia Government Ethics Committee wished to test Hib vaccine under controlled conditions. This necessitated a control vaccine. However, using the ethical comparison set forth by Marcia Angell that research is ethical only when it compares any new interventions with the best available therapy in affluent countries, this would have meant that a placebo control, or even a control vaccine of another type, would be unethical, a position supported by industrialized country reviews.6 Similarly, in a study, also in the Gambia, of 9-valent pneumococcal conjugate vaccine (PCV9), the control vaccine would have to have been the currently approved 7-valent PCV, a product whose serotypes do not represent those most commonly circulating in the Gambia, rather than a placebo, as proposed by the Gambia Government Ethics Committee. Such considerations can (and do) delay trial initiation and cloud interpretation of data. Programmatic decisions. The second type of decision, a programmatic one, is generally made by a national vaccine advisory committee, which recommends whether or not a product should be used based on its safety and efficacy profile, the prevailing disease burden, the epidemiological characteristics of the proposed target population, and the cost of adding the intervention. In the United States, this type of decision is the responsibility of the Advisory Committee on Immunization Practices (ACIP), although other groups, such as the American Academy of Pediatrics (AAP) and the American Academy of Family Physicians (AAFP), might have major input. Many countries have not established such committees. In other countries, decisions taken tend to rely on those published by the ACIP and similar committees in the industrialized world. In the two examples that follow, there are two salient features: a clear difference in risk-benefit ratio in the United States and Europe compared with disease-endemic countries; and little consideration by the industrialized world of the impact of decisions on the rest of the world. Thimerosal. In July 1999, a joint statement by the AAP and the U.S. Public Health Service, following earlier requests by the FDA for vaccine manufacturers to provide detailed information on the thimerosal content of their vaccines, called for a plan to eliminate mercury from vaccines. The statement noted a known and serious risk of not vaccinating against serious disease, relative to the minor and only theoretical risk of vaccinating with products containing thimerosal.7 In June 2000 a further statement—this one also agreed upon by the ACIP and the AAFP—reported progress in achieving this goal but still noted the theoretical risk of thimerosal in vaccines and upheld the goal to remove it based on the feasibility of doing so, progress made to date, and public concern.8 Continuing research has failed to show any ill effects related to thimerosal in vaccines. The WHO, through its Global Advisory Committee on Vaccine Safety, supports its continued use, citing its freedom from adverse effects and the risk to vaccine safety and efficacy occasioned by its removal from some products.9 Moreover, a shift to thimerosal-free products would cause a threat to global vaccine production capacity, with an even greater threat to children’s health if vaccination with these products were compromised. Strong action by the international community averted this situation. Rotavirus vaccine. The FDA approved a vaccine against rotavirus infection, Rotashield, previously manufactured by Wyeth-Lederle, for use in the United States in October 1998. Subsequently recommended for use by the ACIP, the vaccine was administered to more than 600,000 U.S. infants during the first nine months that it was on the market.10 In July 1999 the U.S. Centers for Disease Control and Prevention (CDC) reported fifteen cases of intussusception (a potentially fatal bowel obstruction) potentially associated with use of the vaccine.11 The increased risk of intussusception calculated at that time was judged unacceptable relative to the benefit of the vaccine for U.S. infants, and the ACIP, supported by the AAP and AAFP, withdrew its recommendation.12 The manufacturer voluntarily withdrew the product from the market, although the FDA never withdrew marketing authorization for the vaccine. This decision had dramatic implications for the potential use of this vaccine in the developing world, where the mortality associated with rotavirus infection is much higher than in industrialized countries.13 Following Wyeth-Lederle’s withdrawal of Rotashield, clinical trials in the developing world were either stopped or cancelled. As a result, the risk-benefit relationship for this vaccine was never evaluated in parts of the world where, arguably, the greatest need for a rotavirus vaccine exists.14 The WHO convened a consensus meeting in 2000 that laid out an agenda for future rotavirus research, a key element of which was the need for parallel clinical trials in disease-endemic countries.15 The potential increased risk of intussusception associated with the use of Rotashield continues to be a source of controversy, with estimates ranging from 1 in 2,500 to almost 1 in 30,000.16 Several new rotavirus vaccines are in development, including two late-stage products manufactured by multinational corporations, whose safety is being evaluated in separate programs involving more than 60,000 infants each, a decision supported by FDA statements.17 Thus, decisions by U.S. institutions have delayed access to a vaccine that could greatly decrease early childhood mortality. Product divergence. In recent years there has been wider product divergence between the industrialized world and the developing world, with resulting impact on vaccine supply and price. Thus, in many industrialized countries the immunization schedule includes acellular pertussis vaccines, inactivated polio vaccine (IPV), and conjugated polysaccharide vaccines in single-dose thimerosal-free presentations, while disease-endemic countries use whole-cell pertussis, oral (live) polio vaccine (OPV), BCG, and polysaccharide vaccines in multidose vials with thimerosal. Several factors account for these differences. First, industrialized countries add new products into their immunization schedules that are too costly for the developing world or for which the disease burden is not well documented there. These include mumps, rubella, and, in some cases, varicella and pneumococcal conjugate vaccines. Second, in some industrialized countries, the fear of adverse events or litigation, or both, might have caused a switch to more costly products that could be safer in their context. These include the substitution of acellular pertussis for whole-cell pertussis vaccine, the reduction of thimerosal, and the switch to IPV. Third, sometimes the divergent product performs better. For example, meningitis C conjugate vaccine works more effectively in infants than the polysaccharide products in use in the "Meningitis Belt" in Africa. Finally, the disease burdens of industrialized and disease-endemic countries differ. Thus, for example, yellow fever vaccines and some meningitis vaccine types used in parts of Africa and South America are not routinely used in the industrialized world. These factors also affect the risk-benefit calculation. In the case of products that are divergent because of fear of adverse events or litigation, industrialized countries are choosing to pay a great deal more to lower the risk. The risk-benefit calculation is also affected by additional conditions that influence response to the vaccines. In disease-endemic countries, factors such as anemia, vitamin A deficiency, and pregnancy can modulate responses to both vaccines and diseases.18 Diseases may interact with each other.19 Also, response to BCG vaccine in some people may be lowered because of circulating mycobacteria.20 Implications for the risk-benefit determination. All of these considerations imply a need for risk-benefit determinations based on the disease burden background in the target population. Higher disease burden will mean that high risks with lower benefits could be acceptable in the face of a large number of deaths. Vaccines with lower efficacy or lower safety, or both, might become acceptable in countries with higher disease burden. The need for appropriate frameworks for evaluating the risks and benefits in the context of local or regional rates of infection has been highlighted for vaccines against both HIV and malaria.21 Models of HIV vaccine efficacy have suggested that either preventive or therapeutic vaccines with low or moderate efficacy would have major public health impact, especially if targeted toward populations with the highest risk of HIV infection.22 However, to date there is little precedent for approval of a "partially effective" (less than 50 percent) vaccine in an industrialized country and little guidance for a disease-endemic one to do so.
A double standard? Medical interventions are justified in large part by whether they offer greater prospective benefit than prospective harm. These judgments are often difficult to make because "benefit" and "harm" are difficult to quantify. The prospective benefit and harm of population interventions such as vaccines are even more difficult to quantify, because harm to a few might seem to outweigh a vastly larger benefit to the many. Vaccines that cause serious, tangible net harm to appreciable numbers of healthy people are unlikely to succeed in the marketplace or in popular opinion. These complexities have contributed to an unfortunate and dangerous rigidity in international vaccine practices. The foregoing discussion on rotavirus vaccine is a case in point. Whatever the proper threshold for vaccine risk, the key quantity in question is net predicted risk, which differs hugely in industrialized compared with many disease-endemic countries. Children in the disease-endemic countries have much more to fear from rotavirus than from the rotavirus vaccine. The key element of concern posed by representatives of disease-endemic countries at the 2000 WHO rotavirus conference seemed to be that a decision to approve the vaccine for use in their countries would be viewed there as embracing a double standard.23 In effect, the WHO’s approval would be viewed as being based on the premise that the safety of children in disease-endemic countries was less important than the safety of children in the United States. This reasoning is flawed. If anything, refusal to use the vaccine places a lower value on children in the poorer country, for the original decision was not made in their interests. The net predicted risk from the vaccine in a country where rotavirus is very often lethal is vastly lower than the net predicted risk from not providing the vaccine. But what of the double standard? Sarah Marchand, a philosopher on population health ethics, has proposed a formulation of a single, universal standard for vaccine safety that would yield practices consistent with children’s health: "Every child deserves health care that offers the greatest net benefit."24 This simple standard, if applied around the globe, would sum risks from all sources, including safety. In the United States, where children do not die of rotavirus, the standard might not support populationwide use of the Wyeth-Lederle vaccine; in Bangladesh it would. This principle is set out in a 29 January 2005 New York Times editorial: "Once the vaccine was deemed unsafe in the United States, it was politically impossible for Wyeth to sell it even in countries where it might have caused a handful of deaths while saving 100,000 lives." How should risk-benefit evaluations be made? The guidelines for conducting risk-benefit calculations in research are vague at best. Most guidelines state that a risk-benefit analysis should precede the conduct of research and that this analysis should take full cognizance of harms and benefits beyond the life of the study. Guidelines from the Council for International Organizations of Medical Sciences (CIOMS) for health research classify risk as minimal, low, and high. Minimal risk includes everyday risks such as public transport travel; in medicine it should be no greater than the risk attached to routine medical or psychological examination.25 Although the life experiences of people in different settings will clearly vary, it is unlikely that the level of risk will vary greatly. Benefit can be direct either to the subject or to society. If it is the individual who will benefit, greater risk is usually more acceptable than if society is to be the beneficiary of the research. A guideline may state that however large the benefit, to expose a participant to anything more than minimal risk requires very careful consideration and would rarely be ethical.26
The long-term goal of this action plan is to ensure that health authorities in disease-endemic countries have the capabilities to make vaccine regulatory and policy decisions that are sound from technical, economic, and public health points of view. Whether and how differing risk-benefits ratios can be used for regulatory decision making was central to the discussion that ensued after the withdrawal of Rotashield. From the public health viewpoint, the potential benefits for developing countries outweighed the relatively insignificant risks; from the political perspective, the calculation was less clear. Policymakers in developing countries would face the difficult decision of whether to endorse the use of a product in their country that was deemed not safe enough for children in the United States. A framework for decision making that can guide both international and national policymakers through this difficult debate is outlined below. Information exchange. Define the problem. Successfully addressing the risk-benefit question will require a basis for dialogue that considers a variety of perspectives. This could be achieved through a series of global consultative meetings bringing together regulatory and policy decisionmakers, scientists, ethicists, and vaccine manufacturers from disease-endemic and industrialized countries. Background papers for discussion would clearly state the problems faced by decisionmakers in disease-endemic countries in evaluating the risks and benefits of vaccines. The meeting outcome would be a platform of prioritized issues and proposed actions to address them. This paper, which has highlighted the complexity of the issues, could perhaps serve as a basis for setting the agenda. Develop guidelines. Based on the platform from these meetings, criteria for defining risk and benefit and a decision framework could be developed to ensure a single standard maximizing net benefit. Advocate for clear leadership. Action will require political will on the part of policymakers in endemic countries. The WHO is in the best position to take a leadership role in convening global meetings, reaching consensus on a framework, and developing channels for its application. Process improvement. Provide a stronger science and evidence base for regulatory and policy decisions. For regulatory decisions, a science base refers to the robustness of the evaluation of the product file. Decisions based on theoretical risks may not constructively contribute to public safety.27 Guidelines reflecting the best scientific opinions at the international level can be helpful. For policy decisions, a strong evidence base, including information about disease burden, vaccine impact, cost-effectiveness, and demand, will allow decisionmakers to confidently make committed decisions regarding its value in disease-endemic countries. This need has been recognized by the Global Alliance for Vaccines and Immunization (GAVI), which has provided funding for projects known as Accelerated Development and Introduction Plans that will generate evidence bases for vaccines against rotavirus and Streptococcus pneumoniae infection.28 Ensure parallel vaccine trials. Trials for products intended for use in more than one epidemiological setting should take place in parallel in as many sites as necessary to account for epidemiological differences in disease incidence, circulating strains, and concomitant infections. Some of these are already taking place, funded by product developers, public-private partnerships, or the public sector. Infrastructure strengthening. Continue to strengthen regulatory capacity in disease-endemic countries. The WHO, through its NRA strengthening program and training activities, has made progress in this area, and a recent WHO initiative is a network to improve evaluation of clinical trial data among some developing NRAs.29 Other groups are strengthening their capacity for clinical trial review at the national level. Explore the structuring of advisory committees for regulatory bodies and for vaccine use considerations. Such committees might be based at the regional level, with full participation by representatives of the region. One option is to model such committees after the U.S. ACIP process. Such an approach has been proposed in the area of testing and approving an HIV vaccine, but a process with a broader perspective across a variety of vaccines is needed.30 Strengthen collaboration and harmonization for vaccine regulation. Ways must be found to allow better use of regulatory expertise from industrialized countries in decisions on developing market vaccines. One encouraging action is the ability given to the European Medicines Evaluation Agency to provide technical scientific review of products intended for use primarily in disease-endemic countries, when there is no intention to request marketing authorization for those products in the European Union, the so-called Article 58.31 There is opportunity for the FDA to participate in a similar fashion. The implementation of such a multifaceted pl an could do much to improve access to priority vaccines for disease-endemic countries. International agencies such as the WHO and industrialized countries’ regulatory and policy-making institutions can and should play important roles.
Julie Milstien (milstien{at}medicine.umaryland.edu) is chief, International and Regulatory Affairs, at the Center for Vaccine Development at the University of Maryland School of Medicine in Baltimore and is also an independent consultant in vaccine supply and regulatory issues. Richard Cash is director of the Program on Ethical Issues in International Health Research at the Harvard School of Public Health in Boston, Massachusetts. John Wecker is director of the Rotavirus Vaccine Program at the Program for Appropriate Technology in Health (PATH) in Seattle, Washington. Daniel Wikler is the Mary B. Saltonstall Professor of Population Ethics at the Harvard School of Public Health. John Wecker is director of the Rotavirus Program at the Program for Appropriate Technology in Health (PATH) and is funded in full by the Global Alliance for Vaccines and Immunization (GAVI) and the Vaccine Fund. The authors thank Bruce Gellin, National Vaccine Program Office, U.S. Department of Health and Human Services, for several productive discussions.
| ||||||||||||||||||||||||||||||
Top
Vaccine Policy Decision Making
Ethical Considerations
