Effective Dengue Vaccines: A Pipe Dream?

This is, undoubtedly, the main question dengue researchers have asked themselves for many years. Dengue is one of the fastest-spreading mosquito-borne flaviviral diseases affecting humans (9). The four dengue virus (DENV) serotypes (DENV-1–4) can cause a spectrum of diseases ranging from the self-limiting illness to the potentially lethal severe dengue disease, with bleeding, organ dysfunction, increased vascular permeability, and shock (13). Nowadays, DENV produce >390 million infections and 96 million of them develop at least one sign of severity. Unfortunately, ∼20,000 patients die annually (3). In addition, DENV affects >100 countries in the world and the World Health Organization (WHO) has recognized a 10-fold increase of disease incidence in comparison with the past decade.

To answer this intriguing question, we should analyze many results accumulated during the past 90 years. In 1929, Blanc and Caminopetros reported the first attempt to develop a vaccine against this human pathogen. These scientists inoculated healthy volunteers with sera from DENV-infected individuals and 10 days later they challenged vaccines with a wild-type strain of DENV (4). Although only partial protection was observed in these individuals, the results suggested (a) that a component of the sera (presumably antibodies) can mediate protection and (b) that the development of a vaccine against these viruses is possible. However, in the experiment, the authors did not use a placebo group, limiting the interpretation of their results.

In 1932, Simmons and Reynolds' study supported the idea that a vaccine against dengue could be feasible. Using infected mosquitoes these researchers prepared a homogenate in saline that was treated with phenol and formalin to inactivate the virus. This formulation demonstrated to be safe during preclinical trials in rabbits, pigs, and mice. In addition, the vaccine preparation induced full protection in human volunteers who were challenged 400 days after vaccination (20). In my opinion, this study was the first proof of concept demonstrating we can obtain a vaccine against DENV. However, this evaluation also did not include a placebo group.

More than 20 years after those pioneering experiments, Dr. Sabin isolated the strains Hawaii and New Guinea C of DENV-1 and DENV-2, respectively. These viruses were attenuated by serial passages in brain of newborn mice. With the attenuated strain of DENV-1, Dr. Sabin immunized human volunteers who were challenged 2 or 9 months later with the wild-type strain of DENV-2. As a result, he observed cross-protection depending on the time between the vaccination and the viral challenge (19). These results were the first evidence demonstrating that DENV only induce a short-term protection against heterologous infections. Some years later, Halstead demonstrated that secondary heterologous infections constitute the main risk to develop the severe form of dengue disease (11), one of the main hurdles to develop a dengue vaccine.

In 1963, Wisseman et al. performed the first clinical trial including a placebo group employing as vaccine a mouse neuroadapted viral strain of DENV-1. The vaccine produced an inapparent infection in human subjects eliciting neutralizing antibodies, which began to appear in the second week after vaccination and attained a maximum titer between 3 and 4 weeks after vaccination (27). Vaccine induced partial protection against DENV-3 within 3 weeks after vaccination and persisted at least 85 days (1). This result is in accordance with those described or achieved by Dr. Sabin in 1952.

Epidemiological data collected by Kliks and colleagues in 1988 in babies born from DENV-immune mothers provided the second evidence about the protective role of the humoral immune response. Anti-DENV antibodies transferred from mothers to children protected them against the infection. Nevertheless, this group of scientists observed that after reduction of antibody levels (3–4 months after delivery) due to catabolism, babies became susceptible to the infection and they developed severe dengue during their first natural infection (15). This observation constituted one of the grounds for proposing the phenomenon of antibody-dependent enhancement (ADE) of infection, which is associated with the immunopathogenesis of the diseases.

These antecedents and other studies conducted in animal models a posteriori support the development of a dengue vaccine. However, this vaccine faces several challenges: (a) the four DENV are capable of causing the full spectrum of disease, so in principle a tetravalent formulation is needed, (b) infection induces life-long protection against the infecting serotype but only short-term protection against heterologous infections, (c) secondary infection is strongly associated with severe disease, and (d) partial protection is bad.

After >25 years of research and many clinical trials conducted, the company Sanofi-Pasteur registered a vaccine (Dengvaxia^®) against this pathogen, but the manufacturer has limited its use due to evidence suggesting that it may actually increase the risk of severe dengue under particular circumstances (21). The vaccine cannot be administrated to children younger than 9 years, due to the increased risk of hospitalization observed in this age group (26). Apparently, antibodies generated by the vaccine enhanced the viral infection in this dengue-naive population. This vaccine was developed following a recombinant chimeric live-viral vector approach. It consisted in the substitution of the genes encoding premembrane (prM) and envelope (E) proteins from the attenuated YFV 17D vaccine strain for the prM and E genes of the four DENV. Owing to the absence of DENV nonstructural protein, Dengvaxia elicits a very limited DENV-specific cell-mediated immune response (14).

Other vaccine candidates (TAKEDA's vaccine and NIH's vaccine) under clinical trials have demonstrated to be safe and immunogenic, but their efficacy profile has not been demonstrated (17,25). However, both vaccine candidates include viral regions to develop a cell-mediated immune response.

The re-emergence of viral diseases such as measles, rubella, or poliomyelitis suggests that live-attenuated vaccines against these viruses do not induce a life-long immunity. Probably, booster doses will be required to extend the protection against these diseases. Therefore, any vaccine candidate against DENV under development will probably need booster doses. In fact, Sanofi-Pasteur should evaluate the administration of booster doses to improve the safety profile of Dengvaxia.

In the past years, the protective role of cell-mediated immune response against DENV has taken a crucial role. Several mice experiments have demonstrated that CD8⁺ and even CD4⁺ T cells contribute to protection, controlling the viral disease or reducing viral load in blood and different organs (28,29,31). Although some researchers associated this response with the development of severe dengue during heterologous infection (6,16), a deep study conducted in 2013 demonstrated the correlation between a polyfunctional CD8⁺ T cell response and protection against the disease (23). This study proposes that the human leukocyte antigen (HLA) haplotype defines the quality of the cellular immune response. However, there are certain alleles that during secondary infection inadequately stimulate the CD8⁺ T cells, showing a dysfunctional response (24). Therefore, in the context of ADE of infection, during heterologous viral infections, the T cell dysfunctional response cannot control the viral load, and the infected individual can develop severe dengue. This hypothesis could explain epidemiological data showing that only 3–5% of secondary infections develop the severe form of the disease, despite the existence of cross-reactive antibodies with the potential capacity to induce ADE of infection. However, several studies conducted later have demonstrated the protective role of cross-reactive T cells and its capacity to control the disease in the presence of subprotective antibodies (30,33).

Two important additional concerns related with the development of dengue vaccines were raised by Dejnirattisai and colleagues in 2010 and by Halstead in 2013. The first group of researchers demonstrated that antibodies generated against the structural precursor-membrane protein are highly cross-reactive among the DENV serotypes and, even at high concentrations, do not neutralize infection but potently promote ADE (5). Therefore, any vaccine candidate expressing this protein has a potential disadvantage that is even worse if the vaccine candidate does not contain viral regions to induce a cell-mediated immune response.

In the second study, the author suggests that in accordance with earlier monkey and human studies, a protective immune response to homologous live DENV challenge must be characterized by absence of viremia plus absence of an anamnestic antibody response (solid protection) (10). Nevertheless, DENV infection confers a life-long protection against the infecting serotype, but this immunity could not be sterilizing depending on the interval between the primary infection and the reinfection with the homologous virus. Between 2010 and early 2011, Iquitos, Peru, experienced an unprecedented outbreak of severe dengue caused predominantly by an American/Asian genotype of DENV-2. This epidemic occurred 15 years after the first DENV-2 (American genotype) outbreak in Peru, and authors observed during a population-level analysis that individuals with pre-existing DENV-2-neutralizing antibodies (before the 2010 epidemic) were reinfected by DENV-2 and presented clinically apparent disease (7). These results suggest that incomplete protection against DENV reinfection can occur even after natural infections. Other studies performed in monkeys support that result (2,22). Homotypic antibodies in non-neutralizing conditions can mediate ADE of homologous virus infection, as it has been demonstrated in human and animal models (8,12,32).

All these pieces experimental evidence and results highlight that a dengue vaccine is possible, but unfortunately it is a big challenge. Nobody knows what will be the best antigen for inducing a protective immune response. Neutralizing antibodies generated even against quaternary epitopes can mediate ADE of infection when their concentrations are below an unknown threshold (18). Therefore, several concepts should be changed: (a) neutralizing antibodies could not be necessary, (b) the absence of an anamnestic response is not a correlate of protection, and (c) solid protection against DENV could be a pipe dream.

Vaccine candidates that rely exclusively on cell-mediated immune response could be a potential solution to protect against DENV. This type of vaccines will not completely clear the virus during the infection, but a polyfunctional cell-mediated immune response could be sufficient to reduce viral load during the infection, changing the clinical outcome.