Review

# Accelerating Next-Generation Vaccine Development for Global Disease Prevention

See allHide authors and affiliations

Science  31 May 2013:
Vol. 340, Issue 6136, 1232910
DOI: 10.1126/science.1232910

## Structured Abstract

### Background

Vaccines have provided some of the greatest successes in the history of medicine, including the eradication of smallpox, the near eradication of polio, and the prevention of considerable morbidity and mortality from numerous infectious diseases each year. However, past strategies for vaccine development are unlikely to succeed in the future against major global diseases such as AIDS, tuberculosis, and malaria. For such diseases, the correlates of protection are poorly defined, and the pathogens evade immune detection and/or exhibit extensive genetic variability. Limitations of animal models to predict human immune responses to vaccines, coupled with low success rates for vaccine development compared with biopharmaceuticals, suggest that new paradigms must be implemented for accelerating vaccine development.

Accelerating next-generation vaccine development. Recent advances in vaccine discovery and immune monitoring will enable new human immunology–based clinical research studies to address major gaps in knowledge of vaccine-induced human immune responses and thereby accelerate development of next-generation vaccines.

Recent technological advances in molecular genetics, molecular and cellular immunology, structural biology, bioinformatics, computational biology, nanotechnology, formulation methods, and systems biology are ushering in a new era of vaccine discovery. For example, genomic-based antigen discovery is being exploited for the design of vaccines against multiple bacterial pathogens. Similarly, interrogation of the memory B cell and antibody repertoires from virus-infected subjects has led to the identification of broadly neutralizing antibodies against HIV, influenza, and other viruses, which are now being exploited as tools to design highly conserved epitope-based vaccines. Advances in adjuvant and vector delivery technologies are providing novel approaches for immune potentiation of vaccines, offering new strategies for improving vaccine response rates in neonates and the elderly. However, translation of these advances into vaccines remains impeded by major gaps in our knowledge of human immune responses, including methods to focus immune responses on subdominant protective epitopes, to elicit long-term memory responses, and to drive antibody maturation processes. These gaps can now be addressed given the technological advances described, including the development of approaches to analyze immune responses at the single-cell and systems levels.

### Outlook

Successful development of vaccines against the major global diseases for which vaccines do not currently exist would be transformational for public health, with huge benefits across society. To accelerate next-generation vaccine development, we propose that new human immunology–based clinical research initiatives be established, with the goal of elucidating and more effectively generating vaccine-induced protective immune responses. Collectively, such a "Human Vaccines Project" holds the potential to greatly accelerate the development of next-generation vaccines against major global killers such as AIDS, tuberculosis, malaria, and other infectious diseases; enable more successful vaccine development against allergies, autoimmune diseases, and cancers; and provide a foundation for vaccine development against new and emerging diseases.

## Building Better Vaccines

Vaccines are one of the most effective tools to protect against infectious diseases. Unfortunately, vaccines for diseases with the highest global health burdens, such as HIV, malaria, and tuberculosis, are not yet available. Koff et al. (1232910) review the latest advances in vaccine development and why these particular diseases remain such a challenge. Respiratory syncytial virus (RSV) is a serious cause of morbidity and mortality in infants and young children worldwide. Although a prophylactic antibody is available for children at high risk, a vaccine is much needed. As a potential step toward this goal, McLellan et al. (p. 1113, published online 25 April) solved the cocrystal structure of a neutralizing antibody (D25) bound to the prefusion F protein of RSV. Knowledge of the structure of the prefusion protein should help to guide vaccine design and the development of additional therapeutics.

## Abstract

Vaccines are among the greatest successes in the history of public health. However, past strategies for vaccine development are unlikely to succeed in the future against major global diseases such as AIDS, tuberculosis, and malaria. For such diseases, the correlates of protection are poorly defined and the pathogens evade immune detection and/or exhibit extensive genetic variability. Recent advances have heralded in a new era of vaccine discovery. However, translation of these advances into vaccines remains impeded by lack of understanding of key vaccinology principles in humans. We review these advances toward vaccine discovery and suggest that for accelerating successful vaccine development, new human immunology–based clinical research initiatives be implemented with the goal of elucidating and more effectively generating vaccine-induced protective immune responses.

At the end of the 18th century, Edward Jenner used cowpox-infected materials to immunize against smallpox and introduced the term " vaccine" (1). A century later, Louis Pasteur developed methods for attenuation of bacteria (2), and Salmon and Smith developed methods for inactivation of microorganisms (3). Together, these advances ushered in a new scientific era of vaccinology. Virus propagation in cell culture enabled the development of methods for attenuating viral vaccines (4), leading to a golden age of vaccine development in the second half of the 20th century with the development of several vaccines, including polio, measles, mumps, and rubella (59). By the latter part of the 20th century, most of the vaccines that could be developed by direct mimicry of natural infection with live attenuated or killed/inactivated vaccines had been developed. New technologies, including protein conjugation to capsular polysaccharides and the advent of methods to engineer recombinant DNA, led to the development of vaccines for prevention of bacterial pneumonia and meningitis, hepatitis B, and the recent development of the human papillomavirus (HPV) vaccine (1012). Vaccines have now led to the eradication of smallpox, near eradication of polio, and prevention of untold millions of deaths from infectious diseases each year, and are one of the most effective public health measures available (13). For example, before the introduction of the measles vaccine in the United States, the incidence of measles peaked at nearly 900,000 cases per year, compared with an average of less than 100 cases of measles per year in recent years in the United States (14). Similarly, using metrics to measure cost-effectiveness of vaccines such as disability-adjusted life year (DALY), global vaccination for measles results in US$17 per DALY, one of the most cost-effective health interventions in developing countries (15). Table 1 provides an overview of vaccine-preventable diseases by currently licensed vaccines. Table 1 Major global infections prevented by vaccines. The level of efficacy for the vaccines noted here varies in different populations and regions of the world. View this table: There are several diseases, however, that cause considerable global morbidity and mortality for which vaccines do not currently exist (Table 2). In general, the viruses, bacteria, and parasites for which new vaccines are needed are either much more complex in their pathogenesis, exhibit extensive variability, or have evolved immune evasion mechanisms to thwart the human immune system. For example, there are many cases, such as influenza and dengue viruses, for which immunologic memory induced by natural infection protects against reinfection by homologous serotypes but not by heterologous serotypes (16). Thus, minor changes in the outer glycoproteins from circulating strains of the influenza virus result in the need for annual immunizations against influenza. For viruses such as respiratory syncytial virus (RSV), reinfection with the same virus can occur, although disease is generally less severe with these sequential reinfections (17). For HIV, the hypervariability of the virus coupled with its capacity to integrate in the host genome results in the inability of the host to clear the infection (18). Finally, for pathogens such as cytomegalovirus (CMV), herpes simplex, and Mycobacterium tuberculosis, a carrier state is established with reactivation occurring in situations of immunosuppression (19). Clearly, new vaccine discovery and novel immunization paradigms will likely be required for successful vaccine development against HIV, Mycobacterium tuberculosis, Plasmodium falciparum, hepatitis C (HCV), and other challenging pathogens for which there currently are no licensed vaccines. Table 2 Major global diseases for which vaccines do not currently exist. View this table: Recent technological advances in molecular genetics, molecular and cellular immunology, structural biology, bioinformatics, computational biology, nanotechnology, formulation technologies, and systems biology have heralded in a new era in immunogen design, adjuvant discovery (i.e., agents that enhance immune responses) and immune monitoring. However, translation of these advances into successful vaccines remains substantially impeded by a lack of understanding of key vaccinology principles in humans. This includes the need for greater understanding of disease-specific mechanisms of protective immunity, immune evasion mechanisms, and strategies to drive the immune system toward preferred responses by immunization. Although based on sound scientific principles, currently licensed vaccines have largely been developed empirically, and protection by these vaccines is generally conferred by antigen-specific antibodies, which prevent or reduce infection (20). Viral neutralizing antibodies prevent virus replication by blocking virus binding and entry into cells. For hypervariable viral pathogens such as HIV and hepatitis C, it is likely that broadly neutralizing antibodies (bnAbs) (i.e., those targeting highly conserved regions of these viruses) will be necessary for globally effective vaccines. Many bnAbs against HIV have now been identified from HIV-positive subjects (21), but most of these bnAbs exhibit a high level of somatic hypermutation (22). Moreover, there is currently limited understanding of the optimal immunization strategies in humans necessary to drive the antibody maturation process from germ line to a mature broadly neutralizing antibody (23). Strategies for focusing immune responses on protective epitopes and away from immunodominant decoy epitopes are also undefined, as are the strategies for induction of long-term memory responses. Moreover, most of what we know of human immunology has come from the study of blood, whereas the battle between pathogen and the immune system largely takes place in lymph nodes and other tissues. Finally, little is known of ways to overcome neonatal immaturity and immune senescence in the elderly, important for the development and optimal coverage of next-generation vaccines. Although small-animal and nonhuman primate models serve a critical role in basic science and preclinical vaccine discovery, they are considerably limited when it comes to extrapolation in humans, in large part due to differences in immunogenetics, species specificity of pathogens, and the resident microbiota that influences species-specific responses. Here, we discuss the technological advances that are fueling a new era in vaccine discovery, highlight advances in immune-monitoring technologies and systems biology that offer substantial potential for discovering new biomarkers of protective immunity, and identify the limitations of animal models for screening and prioritizing human vaccines. Based on this analysis, we conclude that new human immunology–based clinical research initiatives with the goal of elucidating and more effectively generating vaccine-induced protective immune responses would greatly accelerate the development of next-generation vaccines against major global killers such as AIDS, tuberculosis, and malaria and provide a foundation for vaccine development against new and emerging diseases. ## The 21st Century Technological Revolution Fueling a New Era in Vaccine Discovery Effective vaccines work by eliciting effector and memory immune responses that confer protection against infection and disease. New technologies have recently been developed that are now fueling a revolution in vaccine discovery, which should help to address vaccine development challenges such as pathogen diversity and immune evasion. These technologies can be divided into three major categories related to antigen discovery, adjuvants and vaccine delivery, and deciphering human immune responses. ### Antigen Discovery Technologies The capacity to sequence whole genomes of microorganisms and use bioinformatics for the design of vaccines is a relatively recent approach to antigen discovery and has been termed "reverse vaccinology" (24). The first pathogen for which reverse vaccinology was attempted was meningococcus B, the cause of 50% of global meningococcal meningitis. With the genome of the bacterium sequenced, more than 600 potential antigens were assessed for antigenicity, resulting in the identification of more than 90 newly detected surface proteins, 30% of which could induce bactericidal antibodies. A subset of these antigens, when formulated as immunogens induced protective immunity in mice, provides proof of concept for ongoing clinical development (25). Reverse vaccinology has now been applied to many other bacterial pathogens as full genome sequencing has advanced, including group B streptococcus, group A streptococcus, Streptococcus pneumoniae, Staphylococcus aureus, and Chlamydia (26). Besides reverse vaccinology, genomic-based antigen discovery has been enhanced by new technologies enabling interrogation of the comprehensive antigenic repertoire using libraries of genetically expressed antigens and screening for immunogenicity of the proteins during infection, termed "antigenome analysis" (27). In addition, advances in mass spectrometry have enabled direct testing of the presence and quantity of antigens on the surface of bacteria (28). Although these reverse vaccinology and antigenome technologies hold considerable promise, particularly for identifying potential antigens for inclusion in vaccines, they remain limited in their capacity to predict which antigens are protective, as recently demonstrated by a phase III trial of an experimental Staphylococcus aureus vaccine (29). Thus, greater efforts on understanding correlates of protection from human natural history studies, coupled with small, iterative clinical trials aimed at driving immune responses toward less immune-dominant but protective epitopes, offers potential for greater use of these antigen discovery technologies. Antigen discovery technologies have also advanced for identification of antigens to be used in vaccine candidates to induce neutralizing antibody responses and for the generation of T cell–based vaccines. For pathogens that are hypervariable, such as HIV reverse engineering of vaccines, the concept that protective antigens could be identified through interrogation of the antibody repertoire from subjects infected with the pathogen (30) and formulated into effective vaccines is now being more fully exploited as a result of recent technological advances in isolating monoclonal antibodies from memory B cells and plasma cells from infected patients, deep sequencing, and bioinformatics. These efforts include (i) identification of subjects with broadly neutralizing antibody serum responses (21); (ii) identification of bnAbs from such subjects by single-cell memory B cell techniques with or without antigen selection and cloning the heavy and light chains into immunoglobulin G vectors (3135); (iii) determination of the structure of the binding sites of such bnAbs at the structural level using crystallographic methods (3135); and (iv) Mimicking the epitopic binding sites of such bnAbs on carrier protein scaffolds or vectors to serve as the basis for immunogens to elicit such bnAbs (32, 33) (Fig. 1). In addition to HIV, bnAbs against HCV (36) and influenza (37) are serving as templates for antigen designs for HCV and universal flu vaccines. The first proof of concept for reverse engineering of vaccines has now been achieved for RSV, where computationally designed immunogens mimicking the binding site for an RSV-neutralizing monoclonal antibody have successfully elicited RSV-specific neutralizing antibodies in monkeys (38). The technologies above highlight advances in vaccine discovery for which antibodies to the pathogen are the principal biomarker of vaccine efficacy. Similarly, there have been major technological advances in the past decade with respect to antigen discovery and evaluation of cell-mediated immune (CMI) responses. CMI responses play an integral role in controlling many acute and chronic diseases, provide critical helper signals for elicitation of antibody responses, and thus are important components of vaccine development strategies against some parasites, intracellular bacteria, and some viral infections. Recent technological advances in CMI assays including enzyme-linked immunosorbent spot (ELISPOT) and intracellular cytokine staining (ICS), mass spectrometry coupled with epitope-identification algorithms (39), tetrameric staining reagents (40), genomics and proteomics of more complex pathogens such as mycobacterium and plasmodium (41, 42), and advances in immune monitoring (see below) have combined to enable advances for T cell–based vaccine antigen discovery. For example, long peptides encompassing conserved regions of HIV have been identified that induce polyfunctional T cells in macaques with breadth superior to single-gene vaccines (43). In addition, computational optimization methods have been applied to the design of "mosaic" proteins, assembled from fragments of natural viral genetic sequences, that provide diversity coverage comparable to that of thousands of separate peptides but are tractable for vaccines (44). These technologies are beginning to be applied both for antigen identification and for assessment of vaccine efficacy. With respect to antigen identification, epitope-specific algorithms aimed at identifying key cellular immunity epitopes have been assessed for vaccinia virus to better understand the protective components of the efficacious smallpox vaccine. Antigens recognized by CD4+ responses differed from those recognized by CD8+ responses, with such data now being applied to reverse vaccinology strategies for development of epitope-specific vaccines (45, 46). Simian immunodeficiency virus (SIV) vaccine efficacy studies with heterologous adenovirus vectors compared in nonhuman primates the level of Gag-specific CMI responses with viral load. Greater control of infection correlated with numbers of Gag-specific epitope responses (47). More recently, using major histocompatibility complex (MHC)–typed monkeys to probe the efficacy of an SIV epitope-based vaccine, it was demonstrated that CMI responses to epitope-specific targets could control SIV infection in nonhuman primates (48). In contrast, studies with malaria immunogens demonstrated that heterologous vectors [chimp Adeno 63 + modified vaccinia Ankara (MVA)] induced robust cellular immunity to specific blood-stage antigens as measured by the magnitude of interferon-gamma ELISPOT responses but had no effect on parasite growth rates in the blood (49). Similarly, a recently completed efficacy trial of MVA85A, a new tuberculosis vaccine in infants previously vaccinated with Bacille Calmette-Guérin, elicited modest CD4+ cellular immune responses as measured by ELISPOT and ICS assays but failed to demonstrate efficacy against tuberculosis (50). ### Adjuvants and Vaccine Vector Delivery Technologies In the past decade, there have been considerable advances in identification of signaling pathways and receptors of the innate immune system and a greater appreciation of the importance of innate immunity in influencing adaptive immune responses (51). Detection of microbes by the innate immune system is largely driven by pattern recognition receptors, such as toll-like receptors (TLRs), that recognize common molecular structures found on microbes. In recent years, there has been a revolution in understanding of the receptors and molecules that drive innate immune responses (5254), and this is now leading to the development and testing of novel vaccine adjuvants. Adjuvants are an important component of vaccine formulations that potentiate immune responses through interaction with one or more TLRs, particularly important for those vaccines composed of proteins that are not highly immunogenic. For example, codelivery of vaccine antigens with pattern recognition receptor agonists may lead to enhanced immune responses to vaccines (5557). This enables more mechanistic-based design of adjuvants for potentiation of humoral and cellular immune responses. By extension, this should help to facilitate the development of next-generation adjuvants to help address vaccine challenges where immune responses may be compromised or less than optimal, such as with immune senescence in the elderly. For example, adjuvant trials in individuals hyporesponive to hepatitis B vaccines (HBV) showed that the addition of CpG to HBV enhanced the kinetics, magnitude, and longevity of the seroprotective response (58). However, there is still a very limited database on improved efficacy in the elderly with adjuvanted vaccines, highlighting the need for additional studies in these populations. Adjuvants currently used in licensed vaccines include aluminum salts, oil in water emulsions, and virosomes and TLR4 agonists such as bacterially derived monophosphoryl lipid A (MPL). In addition, there are a large number of adjuvants currently in development, aimed at boosting CD4+ helper T cell, CD8+ cytotoxic T cell, and humoral immune responses (51). Many emerging and reemerging pathogens, including those responsible for respiratory, gastrointestinal, and sexually transmitted diseases for which vaccines do not currently exist, initiate infection at a mucosal surface, suggesting that induction of mucosal immunity may be necessary or beneficial for prevention and control (59). Vector delivery systems designed to stimulate mucosal and systemic immunity have advanced in recent years, with some already in clinical trials, including adenovirus, paramyxovirus, and some bacterial vectors (60). Viral vectors enable heterologous antigens to be delivered to antigen-processing pathways needed to stimulate human leukocyte antigen (HLA) class I restricted cytotoxic T cell responses, in addition to priming for effective humoral immune responses. Further, replicating viral vectors more closely mimic the attributes of efficacious live attenuated vaccines (61). However, one of the major challenges for vector delivery systems is the need to overcome preexisting immunity specific to the vector. For example, adenovirus type 5 (Ad5) is a potent vector for induction of cell-mediated immune responses, yet the majority of the world's population has been previously exposed to Ad5 or closely related adenoviruses, thus limiting the potential of this vector for vaccine development, particularly for use in the developing world. Strategies to overcome preexisting or nascent antivector immunity are now being assessed, including use of vectors against which humans have low seroprevalence, simian and canine vectors for which preexisting immunity in human populations is negligible, prime-boost regimens, and administration of vector-based vaccines by a mucosal route. Low seroprevalence adenoviruses, such as Ad26, Ad35, and chimpAd63, are currently in clinical trials, and preclinical studies with cytomegalovirus vectors have demonstrated the capacity for overcoming antivector immunity (62). Dendritic cells (DC) are antigen-presenting cells that provide initial surveillance mechanisms for exposure to pathogens, whereupon they undergo rapid maturation and migrate to lymph nodes where induction of immune responses occurs. DCs play a central role in the induction of vaccine-mediated immune responses by providing antigen-specific and costimulatory signals required for T cell activation. Increased understanding of the mechanisms of antigen presentation, guided by elucidation of dendritic cell subsets and their functional plasticity, offer additional opportunities for targeting and potentiation of immune responses by vaccines (63). Targeting dendritic cells in vivo with antibodies to specific DC cell surface receptors is now being explored for vaccines (64). Some viral vectors, such as vesicular stomatitis virus (VSV), target dendritic cells, which may account for the potent immune responses when VSV is used as a vector (65). ### Deciphering Human Immune Responses Along with advances in antigen discovery, adjuvant, and vaccine vector delivery technologies, the past decade has seen a technological revolution in the capacity to analyze immune responses at both the single cell and the systems level, which offers tremendous potential for deciphering human immune responses to vaccines and identification of immune correlates of protection. Advances in systems biology, coupled with recent gains in understanding the pivotal role of innate immunity in augmenting adaptive immune responses, are now being applied to vaccine discovery in what is termed "systems vaccinology" (66). Technological advances with DNA microarrays and high-throughput DNA sequencing, mass spectrometry–powered proteomics, bioinformatics, and computational methods enable data integration that serves as the basis of systems vaccinology (67). Initial studies on yellow fever vaccination in humans have identified early innate signatures that correlate with immunogenicity of the vaccines (68), and similar approaches are now being undertaken for influenza and malaria vaccine development. Interrogation of the antibody repertoire by deep sequencing and bioinformatics has led to the identification of intermediates on the path to broadly neutralizing antibodies that is now helping to guide HIV vaccine development (69) (Fig. 2). Concomitant with advances in systems biology approaches, immune-monitoring technologies have advanced to enable new approaches to characterize and interrogate human immune responses (70). Multiplexed flow cytometric and intracellular cytokine staining assays now allow analyses of cell phenotype, subset identity, activation, and intracellular signaling status and effector activities (7173), and single-cell gene expression analyses allow greater specificity in immune monitoring. Such monitoring should enable greater precision in assessments of correlates of protection for currently licensed vaccines, which will help guide next-generation vaccine development. For example, induction of effector memory responses as determined by flow cytometric analysis is thought now to play an important role in the considerable control of SIV infection conferred by CMV vector-based vaccine candidates (62). Combining these tools with other new methods for analysis of antigen-specific cells using multiplexed tetramer technologies and highly sensitive microarray and sequencing technologies will allow not only for assessment of responses in blood but, because of the sensitivity of these methods, may also provide insights into selected tissue-level responses. This will enable future analyses to be made in the tissues, where the battle between pathogen and host largely takes place, rather than simply relying on measurements from peripheral blood. Taken together, recent progress in technological development for antigen discovery, adjuvant and vector discovery, and technologies aimed at deciphering human immune responses provide the foundation for major advances in vaccine discovery and their application toward accelerating vaccine development against the major global diseases for which vaccines do not currently exist. ## The Importance of Studying Vaccines in Humans: Limitations of Current Animal Models Despite best efforts, in vitro and small-animal models do not effectively recapitulate the dynamics of human immune responses to vaccines. Mouse models, in particular the use of inbred mice, have been extremely effective as a tool for basic immunologists, yet have been largely unsuccessful as models for clinical application (74, 75). For example, inbred mice often have a number of homozygous recessive defects that alter the regulation of immune responses (76). Differences in pattern recognition receptors (e.g., TLR9 expression) may account for considerable differences between humans and rodents in response to microbial stimuli (77, 78). In addition, protocols in small animal models do not adequately reflect human vaccination studies. Many murine studies use intravenous or intraperitoneal injections, whereas human vaccines are generally administered intramuscularly or subcutaneously (79). Different routes of immunization—for example, mucosal versus intramuscular—can change patterns of recognition by dendritic cell subsets, leading to modifications in immune responses (80). In addition, dose and regimen of immunizations are often different between small-animal studies and human clinical trials, which can affect the quality and quantity of priming, effector, and memory responses. As a result of these deficiencies, there is renewed interest in the use of "humanized" mice, which are usually immunodeficient mice that are reconstituted with human hematopoietic stem cells, peripheral blood mononuclear cells, or tissue transplants. Although these models are potentially very promising, they have yet to be validated for predicting human immune responses to licensed vaccines, and human hematopoietic cells developed from stem cell transplantation in immunodeficient mice are not always phenotypically and functionally identical to those that develop in humans (81). Thus, such mice may not fully recapitulate human immune responses, particularly where T cell helper and other effector cell mechanisms are integral for optimization of vaccine-induced immune responses. Although nonhuman primates have played an important role in the development of vaccines for hepatitis B and other diseases, there are additional limitations of nonhuman primate models with respect to human vaccine development, including differences in immunogenetics between macaques and humans, species specificity of some viral vectors being developed as vaccine candidates, and the impact of the microbiome of humans on vaccine evaluation. For example, HIV vaccine developers primarily rely on SIV replication in rhesus macaques as a challenge model for assessing vaccine concepts due to the limited replication of HIV in macaques (82). However, the hypervariability of HIV cannot readily be modeled by SIV because of the limited numbers of referenced strains of SIV. Moreover, differences between MHC restrictions in monkeys and HLA restrictions in humans limits the assessment of epitopes for inclusion in vaccine candidates. In addition, vectors carrying HIV genes such as cytomegalovirus and replication-competent adenovirus type 4 are species-specific for humans. The microbiome of humans consists of the plethora of viruses, bacteria, and parasites that infect and reside in our tissues, contributing a substantial proportion of genetic information to our metagenome and thus affecting susceptibility and resistance to disease, particularly inflammatory diseases such as type 1 diabetes, ulcerative colitis, and Crohn's disease (83). The concept of the microbiome is being applied to personalized medicine but also should be viewed as a limitation when contemplating animal models for human vaccine development (84) . In summary, although murine models have provided important insights into basic immunology, and small-animal models plus nonhuman primates have been important in the development of some currently licensed vaccines, there are considerable limitations of these models in predicting human responses to vaccines. This was demonstrated yet again when simian-human immunodeficiency virus (SHIV) protection studies in monkeys suggested that an Ad5-based HIV vaccine could suppress viral load and control disease, yet human efficacy trials failed to confirm these observations (85, 86). Thus, greater attention needs to be focused on immunogenicity studies in humans aimed at answering specific questions that impede vaccine development and translating this information toward the development of next-generation and more efficacious vaccines against globally important diseases. ## Future Directions: Optimizing Protective Immune Responses in Humans Although licensed vaccines continue to provide tremendous public health benefits, success rates in vaccine development are not optimal and are even worse for the subset of complex pathogens for which variability and immune-evasion mechanisms present additional challenges. For example, in one survey of more than 200 vaccine development projects, success rates for vaccines was only 22%, compared with 40% for biopharmaceuticals (87). In our view, this high rate of failure for vaccines is directly related to (i) lack of information about mechanisms for protective immunity directly applicable to pathogen-specific vaccine development and (ii) lack of understanding of optimal strategies to elicit vaccine-induced protective immune responses in humans and thus inability to effectively predict which immunogens would have a greater chance for success in vaccine efficacy trials. For the major global diseases for which vaccines do not currently exist (Table 2), identification of vaccine-induced protective immune responses will require greater understanding of pathogen-specific mechanism(s) of protective immunity, particularly from human natural history studies and, in some cases, from human challenge models when available, such as malaria. This is important in cases where natural infection by specific pathogens provides protection against subsequent exposure but also in cases where natural infection does not confer such protection. Thus, greater linkages of human disease–specific pathogenesis studies with applied vaccinology studies will be central to accelerating next-generation vaccine development. Moreover, when one factors in the current lack of understanding in humans of (i) how to fine-control the antibody affinity maturation that is likely required to elicit broadly protective responses to highly antigenically variable pathogens, (ii) how to focus immune responses on subdominant yet critical protective epitopes, and (iii) how to elicit long-term central and effector memory responses, it's no surprise that vaccine success rates against complex pathogens such as HIV, mycobacteria, and plasmodium are worse than those against nonvariable, acute infections. However, the confluence of recent technological advances for vaccine discovery, systems biology, and immune monitoring yields a tremendous opportunity for accelerating vaccine development. If harnessed effectively, they can lead to a greater understanding of disease-specific mechanisms of protective immunity, elucidate key principles of vaccinology, and be used to optimize humoral and cellular protective immune responses. Application of this information across the spectrum of disease-specific vaccine development programs will likely shorten the time required for successful development of new and improved vaccines. Clinical trials of vaccines are currently driven by product-specific issues, because sponsors or vaccine developers have preferred to quickly advance candidates from phase I/II safety immunogenicity trials to phase IIb/III efficacy trials rather than prioritize vaccine discovery related issues. Unfortunately, for diseases such as HIV, tuberculosis, malaria, and others, the gap in understanding how best to elicit the requisite humoral and cellular effector and memory responses in humans has slowed the pace of vaccine development. In the absence of a greater understanding of these key vaccinology principles in humans, the vaccine field is left to rely on large and expensive field trials to provide any guidance. Such trials can cost US$50 million to \$100 million or more, can take several years to complete, and often are undertaken with low probabilities of success. This is not to suggest that field trials should be abandoned. It is rather to suggest that, in parallel, considerable resources be directed toward conducting small, iterative, human clinical studies to address key questions currently impeding the development of next-generation candidate vaccines.

In summary, successful development of vaccines against the major global diseases for which vaccines do not currently exist would be transformational for public health, with huge benefits across society. Historical paradigms of empirical product development efforts alone are unlikely to be successful against major global killers that cause substantial morbidity and mortality today. This is due in large part to the current lack of understanding of mechanisms for protective immunity, immune-evasion mechanisms, and immunization strategies on how best to elicit protective immune responses against such pathogens in humans to mimic the best attributes of successful licensed vaccines. To accelerate next-generation vaccine development, we suggest that new human immunology–based clinical research initiatives be established, focused on addressing key questions impeding vaccine development, with the goal of elucidating and more effectively generating vaccine-induced protective immune responses. Collectively, such a "Human Vaccines Project" holds the potential to greatly accelerate the development of next-generation vaccines against major global killers such as AIDS, tuberculosis, malaria, and other infectious diseases; enable more successful vaccine development against allergies, autoimmune diseases, and cancers; and provide a foundation for vaccine development against new and emerging diseases.

## References and Notes

1. Acknowledgments: This work was supported by the network of IAVI donors, including, the U.S. Agency for International Development (USAID) and the Bill & Melinda Gates Foundation (BMGF) (W.C.K and C.R.K); the International AIDS Vaccine Initiative (IAVI) and the National Institutes of Health (NIH) (P.R.J.); the Centers for HIV/AIDS Vaccine Immunology and Immunogen Discovery (CHAVI-ID), with funding from the National Institute of Allergy and Infectious Diseases (NIAID), part of NIH (D.R.B., R.A, and B.D.W). Additionally, S.A.P is a paid scientific advisor and consultant for Sanofi, GlaxoSmithKline, Novartis, Merck, Pfizer, Medimmune, Inovio Pharmaceuticals, Dynavax, and Glycovaxyn, and B.D.W. is a paid scientific advisor and consultant for Merck, Bristol-Myers Squibb, and Globe Immune. The authors also wish to thank S. Glass, L. Gieber, O. Shmaidenko, M. Dees, and B. Hayes for their administrative support.
View Abstract