Dengue human infection model

Abstract

Dengue is an emerging and re-emerging disease of the tropics and sub-tropics. Millions of infections occur annually exacting a significant social, financial, and health care resource toll. Widespread use of a safe and efficacious dengue vaccine in cooperation with strategic vector control is the best hope for reducing the global dengue burden. Despite over 100 y of research exploring dengue immunology, pathogenesis, animal models, and vaccine and drug development there is no licensed vaccine or dengue anti-viral. No correlate of protection or validated animal model of disease has been defined. Experimental human infection with partially attenuated dengue viruses are documented as early as 1902 and have facilitated research efforts resulting in seminal discoveries and observations. It is time to explore re-invigorating the dengue human infection model to support dengue vaccine development.

Keywords: :

• dengue
• vaccine
• experimental
• human
• infection
• challenge

The world needs a dengue vaccine. Unfortunately, diverse efforts and approaches spanning more than 70 y have failed to achieve this goal.¹ As a neglected tropical disease, dengue and the dengue vaccine field have not received the consistent financial or political support benefitting other infectious diseases. However, fundamental scientific challenges have also contributed to slowing vaccine development efforts. It is time to explore re-establishing the dengue human infection model (DHIM) to support dengue medical countermeasure development and advance the field’s understanding of this disease’s unique immuno-pathogenesis.

Dengue is the world’s most rapidly expanding and important arboviral disease.² Millions of people are infected, millions suffer symptomatic disease requiring medical care, and tens of thousands die every year from endemic and epidemic transmission.³(WHO, 2013. Dengue and severe dengue, Fact sheet N°117. Available at: www.who.int/mediacentre/factsheets/fs117/en/index.html, accessed February, 2013.) Dengue is a leading cause of febrile illness in the returning traveler and has impacted military operations since World War II.^4-13 Although mortality is relatively low compared with other infectious diseases of the developing world, dengue exacts a significant social, financial, and health care resource toll.^14-17

There is no licensed vaccine or therapeutic to prevent or treat DENV infection or disease. As such, many endemic countries execute costly prophylactic or reactionary vector control measures, often without noticeable impact. Population growth and urbanization, changing ecologic conditions favoring expanding Aedes mosquito habitats, and the increasing ease and access to regional and international travel will continue to drive the worsening global dengue burden.¹⁸

In 2012 there was expectant optimism surrounding the results of the world’s first dengue vaccine efficacy trial. Optimism gave way to confusion and disappointment as the vaccine’s sponsor published in The Lancet an overall per-protocol efficacy of 30.2%. The trial occurred in a population of Thai children highly immunologically primed to Japanese encephalitis virus and/or dengue, a condition many thought would improve the vaccine’s performance. Efficacy across the dengue virus (DENV) types was unequal; efficacy against disease following DENV-2 infection was the lowest (9.2%). Neutralizing antibody responses in vaccine recipients compared with control were robust and balanced. There was no evidence symptomatic breakthrough infections in the vaccine group were more severe than in controls. Numerous hypotheses have been proposed to explain the study findings and the apparent disconnect between immunogenicity and clinical efficacy. Unfortunately, the size and blood sample collection schedule of the per-protocol immunogenicity volunteer subset is unlikely to support a comprehensive retrospective forensic analysis.¹⁹

The limited success of this first efficacy trial has highlighted a number of questions for the field:

• Is the importance of neutralizing antibody in question as it relates to its role in the protective immune response following natural infection or it’s potential to be validated as an immune correlate of protection?

• Is development of DENV type-specific cellular immune responses against the DENV non-structural proteins a requirement for protection?

• Have concerns increased regarding the limitations of the plaque reduction neutralization test (PRNT) assay platform and its ability to measure vaccine immunogenicity?

• Do dengue vaccine developers need to increase their consideration of DENV diversity and evolution when developing vaccine candidates?

• Is it possible population-level genetic background and pre-existing antibody levels to dengue and/or non-dengue flaviviruses could significantly impact vaccine candidate performance?

One method of answering these questions would be a prospective cohort study measuring clinical endpoints. Scheduled cohort blood sample collections would allow retrospective analyses of the relationships between interesting clinical endpoints (e.g., any dengue, any clinical severity, caused by any infecting DENV type) and immune profiles (i.e., neutralizing antibody titer) measured in pre-illness samples. These studies are expensive and require enrollment of large sample sizes to address clinical attack rates of 1–2% per year.^20,21 Risk of capturing sufficient clinical disease is risky in areas with focal and unpredictable transmission dynamics.^22,23 The more narrow the study questions (i.e., study of severe dengue, study of disease caused by a specific DENV type, etc.) the larger the study.

Another option for exploring the relationship between immune profiles and protection would be to use a well-characterized dengue disease animal model. Although significant progress has been made in this area, gaps remain and vaccine developers are reluctant to base important development decisions (i.e., investing in a clinical endpoint study) on a candidate’s efficacy demonstrated in a mouse model.²⁴ Non-human primates (NHP) develop viremia and a neutralizing antibody response following exposure but usually do not become ill.^25,26 Therefore, vaccine development decisions are guided by extrapolating efficacy data from NHP studies (i.e., prevention of viremia in vaccinated vs. controls following challenge) and assumptions about the association between neutralizing antibody profiles (i.e., magnitude, balance) observed in phase 1 and 2 clinical studies and the candidate’s potential for clinical benefit.²⁷

Response to The Lancet publication should be measured and final judgment on vaccine performance reserved until phase 3 data from a broader population and geographic area is available. It is prudent, however, for dengue vaccine developers to believe their development programs may be at increased risk if efficacy in a NHP dengue infection model and neutralizing antibody are not predictive or associated with human clinical efficacy in any meaningful way or in any way which can be measured. For these reasons, it is time to explore re-establishing a DHIM.

For decades, controlled human infection (challenge) studies have been used to develop vaccine candidates and study infectious disease immunology.^28-33 The ethics of intentionally infecting healthy volunteers has been vigorously debated and much has been written on the topic.^34-36 A discussion of these issues is beyond the scope of these comments. However, one unique challenge of a DHIM, which many believe differentiates it from other controlled infection scenarios, is the absence of any specific anti-dengue therapeutic (i.e., anti-viral). Although specific treatment algorithms do exist and in experienced hands work very well to prevent morbidity and mortality, many find the absence of a specific anti-dengue therapeutic concerning. DHIM developers and end-users will be required by regulatory agencies and institutional review boards to calculate and articulate the risk of a bad outcome (i.e., severe dengue, dengue hemorrhagic fever) following infection with a slightly attenuated DENV. In-depth descriptions of measures to reduce volunteer risk and clinical management guidelines will need to accompany protocol submission packets.

Descriptions of controlled dengue human infection experiments can be found in the literature as early as 1902 and as recent as 2013.^1,37-47 According to a literature review, more than 670 volunteers have been experimentally infected using both needle and mosquito virus delivery. Early foundational understandings about dengue were produced using human infection experiments. Discoveries and observations included: the viral etiology of dengue, virus transmission mechanisms, viral incubation period in mosquito and man, period of infectivity in mosquito and man, competence of Aedes species to transmit virus and the incompetence of Culex species, clinical and clinical laboratory features of uncomplicated dengue illness, development of homologous immunity following infection, the existence of multiple DENV types, demonstration of anti-dengue virus neutralizing antibodies, and the utility of a DHIM in assessing vaccine candidates. More recent studies have reported ultrasonographic findings, characteristics of cellular immune responses, and clinical and clinical laboratory findings following experimental infection of flavivirus naïve and previously vaccinated (experimental candidate) individuals.^48-50

A well-characterized DHIM would be a valuable research tool. Vaccine and drug development pathways, immunologic investigations of pathogenesis, exploring immune correlates of protection, and augmenting regulatory strategies for licensing and labeling products would all benefit. Vaccinologists, drug developers, and immunologists discussed these concepts at a Walter Reed Army Institute of Research (WRAIR) and National Institutes of Health (NIH) co-sponsored, “Dengue Human Infection Model,” workshop in 2011 and a recent Wellcome Trust event entitled, “Controlled Human Infection Studies in the Development of Vaccines and Therapeutics.” Comments here will focus on vaccine development.

A DHIM could support various elements of the vaccine clinical development life cycle. Vaccine developers and scientists could test their candidate’s safety and potential for clinical efficacy across a number of conditions and scenarios. Endemic countries introducing a dengue vaccine into their national immunization program will undoubtedly vaccinate numerous previously infected individuals. Therefore, a dengue vaccine will need to establish, early in development, safety in dengue primed individuals.⁵¹ Initial explorations of safety in dengue primed individuals typically require execution of phase 1 or 2 studies in dengue endemic countries.^52,53 Preparing for and conducting these studies is expensive and painstakingly slow. Depending on the region, study volunteers may also be primed to other, non-dengue flaviviruses (i.e., Yellow fever, Japanese encephalitis) making data interpretation difficult. A DHIM would allow developers to create dengue primed individuals and complete preliminary safety testing on a smaller scale and in very controlled and lower risk environments. The impact of dengue priming on safety as a function of time (i.e., recent vs. remote priming) could also be evaluated. Priming to different DENV types would be possible. In-depth clinical and immunologic characterizations of volunteer reactions to vaccination could generate early safety benchmarks and promote hypotheses for observed outcomes (i.e., reactogenicity associated with pro-inflammatory cellular immune responses).

A DHIM could also support early vaccine formulation decisions including antigen selection, antigen concentration, balancing antigen input, and adjuvant selection and dosing. Challenging vaccine recipients across different treatment groups and candidate formulations would supplement NHP and human immunogenicity data to allow more informed decision-making earlier in development. If a DHIM for every DENV type was available, developers could probe DENV-specific vaccine performance and target re-formulation or re-derivation efforts. Improving vaccine performance or other aspects of the target product profile (TPP) with an adjuvant could also be explored.

One of the most valuable uses of a DHIM would be supporting the discovery of an immune correlate and surrogate of protection. Correlating pre-infection immune profiles with post-experimental infection clinical outcomes would allow scientists to explore how humoral and/or cellular immunity readouts may be associated with protection. It is possible vaccine-induced neutralizing antibodies or interferon gamma responses may be associated with disease attenuation or prevention (i.e., correlate). Quantitative protective thresholds or breakpoints for individual endpoints may also exist (i.e., surrogate). Correlates and surrogates would be viewed in the context of the specific assay platform and executing laboratory (i.e., vaccine developer “X” using a microneutralization platform) and determined in a DENV type-specific manner. The timeframe between vaccination and sample collection (i.e., 1 mo post-completion of the vaccine regimen, 6 mo post-completion, etc.) would be considered. DHIM-derived correlates and surrogates would allow developers to establish target benchmarks for vaccine candidate performance in early phase 1 and 2 studies. Candidates not meeting the threshold could be down-selected or re-derived. Recognizing surrogate threshold may not hold true following infection with an un-attenuated wild type DENV captured during an efficacy trial (i.e., DHIM threshold too low), DHIM derived correlates and surrogates would need to be validated by associating them with actual efficacy and immunogenicity endpoints measured during a field trial.

The above represents an incomplete list of potential uses for a DHIM in dengue vaccine development. Uses related to exploring basic immunology, clinical pathology, drug development, and diagnostics also exist. Augmenting regulatory strategy for licensing vaccines or drugs in non-endemic countries requires further discussion with national regulatory authorities.

The DHIM would bring value to vaccine development efforts because of its high “negative predictive value.” A vaccine candidate capable of protecting a volunteer following infection with a slightly attenuated DHIM viral strain may be protective in a field efficacy study with natural wild-type viruses. On the other hand, there is little chance a vaccine candidate would be protective in the field if it was not protective in the context of a DHIM study. The need for a dengue vaccine is too great and efforts to develop a vaccine have proceeded unsuccessfully for too long to not explore a DHIM’s utility and potential concerns.

Abbreviations:
DENV	=	dengue virus
DHIM	=	dengue human infection model
NIH	=	National Institutes of Health
NHP	=	non-human primate
PRNT	=	plaque reduction neutralization test
TPP	=	target product profile
WRAIR	=	Walter Reed Army Institute of Research

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Disclaimer

The opinions or assertions contained herein are the private views of the author and are not to be construed as reflecting the official views of the United States Army or the United States Department of Defense.