Resurgence of Interest in Eastern Equine Encephalitis Virus Vaccine Development

Abstract

Eastern equine encephalitis virus (EEEV; Family Togaviridae), is an endemic pathogen first isolated in 1933 with distribution primarily in the eastern US and Canada. The virus has caused periodic outbreaks in both humans and equines along the eastern seaboard and through the southern coastal states. While the outbreaks caused by EEEV have been sporadic and varied geographically since the discovery of the virus, it has continued to expand its range moving into the Midwest states as well. Additionally, one of the largest outbreaks was recorded in 2019 prompting concerns that outbreaks were becoming larger and more frequent. Because the virus can cause serious disease and because it is transmissible by both mosquitoes and aerosol, there has been renewed interest in identifying potential options for vaccines. Currently, there are no licensed vaccines and control relies completely on the use of personal protective measures and integrated vector control which have limited effectiveness for the EEEV vectors. Several vaccine candidates are currently being developed; this review will describe the multiple options under consideration for future development and assess their relative advantages and disadvantages.

Introduction: History of Eastern Equine Encephalitis

While the earliest description of probable Eastern equine encephalitis (EEE) infections occurred in 1831 with reports of equines with neurological disease, eastern equine encephalitis virus (EEEV) was first isolated during outbreaks of equine neurological disease in 1933–1934 in Virginia, Maryland, and Delaware (Morris 1988). Equine outbreaks varied in size but could be quite substantial typically with a significant mortality rate. The virus was also determined to affect several species of avians including ring-necked pheasants, sparrows, pigeons, partridges, emus, and ostriches among others; while some avians are associated with enzootic maintenance some species of birds have been found to suffer neurological effects of EEEV (Tyzzer et al. 1938, Kissling et al. 1954, Morris 1988).

Humans were first confirmed to be affected by the virus in 1938 when the virus was isolated from a human brain (Fothergill et al. 1938). A handful of human cases have been reported in most years since this initial event with a median of eight neuroinvasive disease cases reported each year between 1997 and 2007 and 2003–2016 (Lindsey et al. 2018). However, there have been unusually active years with a much higher number of human cases: in 1959, the second-largest human outbreak was reported in New Jersey with 32 cases (Goldfield and Sussman 1968) while in 2019, a total of 38 cases occurred (Fig. 1). Significantly, the mortality rate can be quite high in those presenting with neurological disease; of the cases reported between 2003 and 2016, 41% of the affected patients died (Lindsey et al. 2018).

Fig. 1.

Cases of eastern equine encephalitis (neuroinvasive diseases) reported by year. https://www.cdc.gov/easternequineencephalitis/tech/epi.html

Biology of EEEV

Like all alphaviruses, EEEV has a genome consisting of a single strand of positive-sense RNA that encodes four non-structural proteins (nsP1–4) in the 5’ two-thirds of the genome. There are three structural proteins plus three peptides (capsid [C], envelope glycoprotein 1 [E1], envelope glycoprotein 2 [E2], 6K, transframe [TF], and peptide E3) in the 3’ one-third of the genome (Fig. 2) which are generated as a polyprotein from a subgenomic messenger RNA (Snyder et al. 2013). During the virus maturation process, the capsid associates with the genomic RNA to form nucleocapsids; these are coated with the surface glycoproteins (E1 and E2) at the cell surface during the budding process. The non-structural proteins are not packaged in the final virions so the humoral immune response primarily targets the predominant structural proteins present on the virion surface (E1 and E2). These surface proteins are also the target of most vaccine and antiviral strategies as will be discussed below.

Fig. 2.

Genome organization of EEEV. The figure denotes both the order and presence of the structural and non-structural proteins encoded in the genome. * Element replaced by IRES for LAV candidate; # Region utilized in replicon particle candidates; % Genes expressing antigenic epitopes used in vectored, chimeric, and DNA candidates; ^ Gene component used in subunit candidate.

Early EEEV Vaccine Development

Vaccines have been in development for over 80 yr with renewed interest in the past decade. The earliest EEEV vaccine work described was in 1938 when chick embryo vaccines were used to protect equines (Lowenthal et al. 1961). Equine vaccines are commercially available, so this review focuses on human vaccine products in development. Human EEEV vaccine research was initiated in the 1940s by the Department of Defense (DoD) to protect military personnel (Reisler et al. 2012). In the 1960s, the EEE vaccine program at the United States Army Medical Research Institute (USAMRIID) examined formalin-inactivated chick embryo cell cultures of the PE-6 strain of EEEV. When testing this product in at-risk laboratory workers, they found that 2 doses provided neutralizing (NT) antibody (Ab) titers in 93% of the subjects (Bartelloni et al. 1970). Further evaluation of inactivated PE-6 continued through the 1980s leading to the designation of the vaccine candidate lot as an Investigational New Drug (TSI-GSD 104). This vaccine was administered as a 2-dose regimen given subcutaneously 1 mo apart (Reisler et al. 2012). In total, 60% of subjects in this protocol generated an 80% plaque reduction neutralization titer (PRNT₈₀) titer of at least 40 after the 2-dose primary series. An intradermal booster dose at 6 mo resulted in an 84% response rate (Maryam et al. 2019). These neutralizing antibodies were found to persist for approximately 2 yr, but additional booster doses were required to maintain long-term antibody levels. However, the vaccine was found to have few side effects as might be expected from a killed vaccine product.

Current Vaccine Options

In the early 2000s, there was a renewed interest in developing vaccines to protect against EEE disease as well as the other encephalitic alphaviruses. A variety of research approaches focused on novel vaccinology platforms were used to develop a product that would generate strong and lasting immunity, be cross-protective, and require minimal and straightforward production needs. Options including virus-like particles (VLPs), subunit vaccines, vectored/chimeric vaccines, nucleic acid vaccines, and live attenuated vaccines have all been explored as possibilities (Table 1). These products will be reviewed below with the advantages and disadvantages of each described.

Table 1.

EEEV vaccine candidates under development and evaluation parameters of each

Vaccine type	EEEV element/source (genotype)	Model system	Immunization			Challenge			Humoral immune response	References
			Dose	Route	Schedule	Dose	Route	Strain (Genotype)
VLP	Lentivirus expressed C – E1, PE-6 strain	BALB/c mice, 6–8 wk cynomolgus macaques, 3.2–8.4 kg	5 μg VLP 20 μg VLP	i.m. i.m.	0,21 d 0, 28 d	8.9 × 103 pfu, 56 dpi 108 pfu, 56 dpi	aerosol aerosol	FL93–939 FL93–939	IC50 5240 IC50 >10,000	Ko et al. 2019
Subunit	Baculovirus expressed E1, WEEV McMillan strain	CD-1 mice, 4–6 wk	10 μg/ 200 ul dose	s.c.	0,14 d	1–5 × 104 pfu, 9 wk pi	s.c., i.n.	FL93–939	ND	Phillips et al. 2014
Inactivated	CVEV1219 strain (recombinant), based on FL91–4679	BALB/c mice, 6–8 wk	0.1,1, 3, 5 μg	S.C., i.m., i.n.	0, 28 d	0.1–5 μg (>100 LD50), 28 dpi	aerosol	FL93–939	s.c.: NT80 <1000 i.m.: NT80 <1000 i.n.: NT80 <100	Honnold 2014
Chimeric virus	MVA [vaccinia] E3 – E1 FL93–939	BALB/c mice, 16–18 g	1 × 108 TCID50	s.c.	0, 28 d	5 × 103−1 × 104 pfu, 14 d post-boost	i.n.	PE6	NT50 log 1.88–2.88	Hu et al. 2018
Chimeric virus	ISFV [vesiculovirus] E3-E1, FL93–939	CD1 mice, 4–6 wk	107 pfu	i.m.	Single dose	104 pfu, 4 wk 105 pfu	s.c., i.p.	FL93–939	80% with NT80 ≥20 at 14 d post-vaccination	Nasar et al. 2017
Chimeric virus	SIN/NAEEEV, C – E1, FL93	cynomolgus macaques, 3–6 kg	5.0 log10 pfu	s.c.	Single dose	45–66 d post-vaccination	aerosol	FL93–939	83% with NT80 ≥20 at 12, 24, or 45 dpi	Roy et al. 2013
Chimeric virus	EILV/EEEV C-E1, FL93–939	CD-1 mice, adult	108 pfu	s.c.	Single dose	105 pfu, 70 dpi	i.p.	FL93	NT80 ≥80 at 28 dpi	Erasmus et al. 2018
Live attenuated	EEEV/IRES FL93–939	Swiss NIH mice, 3–4 wk	104 pfu	s.c.	single dose	103 pfu; 4 wk pi	i.p.	FL93–939	NT80 160–640	Pandya et al. 2012
nucleic acid (replicon, virus-like replicon particles)	nsl-4; E3-E1; FL91–4679	BALB/c mice; 6–8 wk cynomolgus macaques	107 IU	s.c	0, 28 d	105 pfu; various times post vaccination 108 pfu; 60d post prime	aerosol	FL91–4679	NT80 640–11763 Mean NT80 163 at 56 d post vaccination	Reed et al. 2014
nucleic acid	E3-E1; FL91–4679	BALB/c mice, 6–8 wk	5ug plasmid	Electrode array, i.m.	0, 21, 42 d 0, 28, 230 d	105 pfu (~3000 LD50)	Aerosol ND	FL91–4679	Mean log10 NT80>2 at d42 pi	Dupuy et al. 2018
		New Zealand white rabbits; 3–3.5 kg	0.5 mg plasmid						Mean log10 NT80 267 at d230 pi

Pi, post immunization; NT, neutralizing titer; IC, inhibitory concentration; d, days; wk, weeks; y, years; ND, not done or not discussed; iu, infectious units; s.c., subcutaneous; i.p., intraperitoneal; i.v., intravenous; i.d., intradermal; i.n., intranasal; LD, lethal dose.

Next Generation Inactivated Vaccine

One of the most classic approaches to the development of a vaccine involves the inactivation of a live virus, typically through heat or chemical means. However, this approach usually involves the growth of virulent virus and has the risk of the final product not being completely inactivated which could result in severe disease. The inactivation of an attenuated strain decreases this risk but still involves the growth of live viruses and requires methods to ensure complete inactivation. Additionally, inactivation methods can alter the physical structure of viral epitopes which could result in a poorly immunogenic product. One recent attempt to develop a new generation of inactivated EEEV vaccine examined the use of a variety of inactivating agents including formalin, INA, and gamma radiation (Honnold et al. 2014). An attenuated, recombinant EEEV strain, CVEV1219, was subjected to treatment with each of these reagents and then evaluated for both immunogenicity and protective capacity in a mouse model. Because EEEV can be transmitted by aerosol, this was the challenge route evaluated in the efficacy studies.

The initial vaccination and challenge studies used a variety of inactivation methods, a range of doses, different routes of inoculation, and single or prime/boost strategies. If the challenged animal survived, there were only infrequent clinical signs observed; however, most of these symptomatic animals did not survive. These findings were not affected by dose, route, or inactivation method. In general, two vaccine doses were more protective than single doses and formalin inactivation was more protective than other inactivation methods. In analyzing the immune response, the level of neutralizing antibodies was found to be the most protective immune element against lethal aerosol challenges. However, cell-mediated immunity was not evaluated in this work (nor for most of the current candidates) to determine its role in protection. Continued efforts to find a combination of inactivation methods that elicits an optimal immune response are still required as these initial studies found significant variability among all the groups tested (Honnold et al. 2014, Wolfe et al. 2014).

Live Attenuated Virus

One straightforward approach to develop a strong safety profile in a vaccine candidate is to attenuate a wild-type, virulent virus. This can be accomplished by the repeated passage of the virus in cell culture (as was done for the alphavirus, Venezuelan equine encephalitis virus, to generate the TC-83 vaccine) (Berge et al. 1961). However, with the advancement of reverse genetic methods, engineering in select mutations to achieve an attenuated phenotype is increasingly common. Live attenuated vaccines (LAVs) typically are more immunogenic than inactivated products and manufacturing can be easier since no inactivation steps (or verification of inactivation) are required. However, the potential to revert to a virulent form is a significant concern for many LAV’s.

One potential EEEV live-attenuated candidate was generated by replacing the subgenomic promoter responsible for producing the structural polyprotein with an internal ribosome entry site (IRES) derived from the encephalomyocarditis virus (EMCV). Not only did incorporating this single element attenuate the virus in cell culture (Pandya et al. 2012), it has the added safety advantage of inhibiting replication in mosquito cells (as the CMV IRES element is not functional in invertebrates) (Finkelstein et al. 1999). This safety feature means that mosquitoes ingesting blood from a vaccinated individual could not replicate nor transmit the vaccine virus. When this approach was used with the alphavirus chikungunya virus, the resulting product was optimally attenuated yet immunogenic (Plante et al. 2011) and could not be transmitted by mosquitoes. Subcutaneous inoculation of the EEEV/IRES virus in 3–4 wk old Swiss NIH mice resulted in the detection of neutralizing antibodies in all immunized mice with no detectable viremia (Pandya et al. 2012). Upon challenge with a virulent virus, none of the immunized mice succumbed and no virus was detected in brain tissue after sacrifice. As expected, the EEEV/IRES virus could not be detected in mosquitoes that had been intrathoracically inoculated. While most LAV vaccine candidates must be monitored to ensure compensatory or adaptive mutations that could lead to reversion of virulence do not develop, that is less of a concern when an entire promoter element is replaced. Additional studies to evaluate if this candidate is protective against aerosol challenge remain to be performed.

Virus-Like Particles

One potential new EEE vaccine involves the assembly of VLP’s. Like other alphaviral VLP’s, the particle is generated from the viral structural genes, C-E3-E2–6K/TF-E1, but no nucleic acid is incorporated within the intact virion structure. This lack of nucleic acid means the VLP approach has a greater safety profile than live-attenuated options as no live virus is present nor could be generated even though the VLP is physically identical to an infectious virus. Additional favorable attributes include ease of production and the capacity to rapidly generate large stocks. However, because it is functionally, a “killed” product, concerns of low immunogenicity may have some merit.

One EEEV VLP vaccine candidate has been shown to be highly immunogenic in BALB/c mice generating significant levels of NT antibodies after 2 doses (Ko et al. 2019). This vaccine was further evaluated in a cynomolgus macaque challenge model. Animals were inoculated twice with VLP’s intramuscularly four weeks apart after which the VLP’s again elicited strong neutralizing antibodies. Vaccinated animals were then challenged with wild-type viruses by the aerosol route and monitored for signs of illness. All animals receiving the EEEV VLP were fully protected from infection and showed no signs of pathological lesions upon examination of brain tissue. In contrast, the majority of the sham vaccinated animals demonstrated signs of viral encephalitis and succumbed to infection within a week. While this was evaluated in an aerosol challenge model and may also suggest protection by other routes of viral transmission, the more conventional infection route (s.c., by mosquito bite) should also still be examined.

Chimeric Vaccines

One other recently explored option for vaccine development is to use a virus–vector system which incorporates genetic elements of the target agent that are expressed from a vector or shuttle virus resulting in a strong antigen response. These vectored vaccines have the advantage of continuous replication which may result in more robust immune responses. Additionally, because they contain avirulent or attenuated backbones, they frequently have increased safety profiles. A number of different vector platforms have been developed for alphavirus vaccines (Powers 2018); these that have been developed to target EEEV are described below.

Alphavirus Based Chimeras

A number of alphavirus candidate vaccines are based on combining the genomes of two distinct alphaviruses. One EEEV chimeric alphavirus candidate is based on the open reading frame (ORF) of the structural proteins of EEEV combined with the non-structural coding region of Sindbis virus (SINV). Typically, chimeric alphaviruses containing the replication machinery of one alphavirus and the structural genes of another are highly attenuated (Paessler et al. 2003). However, because they are also live vaccines, they continue to replicate and express the antigenic proteins of interest. This continued expression results in a robust immune response that often is not generated from killed vaccine products. The SIN/EEEV chimera was shown to generate no viremia and high levels of neutralizing antibody in a mouse model and completely protected mice from a lethal intraperitoneal EEEV challenge (Wang et al. 2007). The SIN/EEEV chimera was further tested in cynomolgus macaques since this is an animal model that most closely mimics human alphaviral neurotropic disease (Roy et al. 2013). Eighty-three percent of the macaques that were vaccinated with the SIN/EEEV chimera developed an 80% plaque reduction neutralization titer and five of the six animals developed IgM antibodies. When the macaques were challenged by aerosol with a lethal dose of wild-type EEEV, 82% of the animals survived compared with no survival in the sham-vaccinated group. Additionally, no lesions were found in the brains of protected animals upon histological examination. These studies demonstrated protection by target virus structural protein expression in both subcutaneous and aerosol transmission routes, a feature that would be essential for any EEE vaccine.

A further safety development in chimeric alphavirus vaccine development involves the use of the insect-only alphavirus, Eilat virus (EILV), as the backbone and the structural genes of EEEV. Because EILV is an insect-only alphavirus, it lacks the capacity to replicate in vertebrate cells thus functionally serving as a killed vaccine. To determine the safety profile of this product, 10⁸ plaque forming units (pfu) of the EILV/EEEV was injected directly into the cranium of 7 d old CD-1 mice (Erasmus et al. 2018). All animals survived with no detectable neurological lesions. Subcutaneous vaccination of mice resulted in 60% seroconversion by 6 d postimmunization with PRNT₈₀ titers of 20–40. All animals seroconverted with titers of 80–640 by 28 d postimmunization. Intraperitoneal challenge of immunized mice resulted in 100% protection from mortality. This virus is a strong EEEV vaccine candidate as it has elevated safety, is rapidly immunogenic, and is effective with only a single dose. However, because this is functionally an inactivated candidate, the long-term immunogenicity must be determined, and the effectiveness of the product needs to be evaluated in an aerosol challenge model and with NHP’s to determine whether further development is warranted.

Vaccinia Based Chimeras

Modified vaccinia virus Ankara (MVA) has been developed as a vector vaccine system that has been widely developed for a number of agents (Sutter and Staib 2003, Gómez et al. 2011). One advantage of this system is the safety profile; the virus is highly attenuated as productive viral assembly does not occur in most mammalian cells (Volz and Sutter 2013, Hu et al. 2018). It has been shown to readily express foreign genes and induce both humoral and cell-mediated immunity. An MVA system was developed that included the glycoprotein cassette of EEEV (E3 through E1; strain FL93–939) (Hu et al. 2018). BALB/c mice were inoculated subcutaneously with 10⁸ of 50% tissue culture infectious doses (TCID₅₀) of the MVA EEEV virus (MVA-BN-E) in 2 doses given 28 d apart. These mice generated neutralizing antibodies by day 14 with increased NT antibody titers detected at 41 d. When mice were challenged with wild-type virus, no disease symptoms were detected in vaccinated mice and all animals were protected from a fatal disease. While the NT antibodies were presumed to provide this protection, assessment of the role of T cells and innate immunity were not examined in this study.

Vesiculovirus Based Chimeras

An additional chimeric virus system evaluated as a possible EEEV vaccine utilizes a vesiculovirus backbone based on Isfahan virus (ISFV) (Nasar et al. 2017). Previously, vesicular stomatitis virus (VSV) expression vectors have been used in a number of vaccine studies by generating pseudotyped viruses with genes from several different pathogens (Roberts et al. 1999). The success of using VSV led investigators to examine additional vesiculovirus vectors that are distinct from VSV to avoid any problems arising from neutralization of the vector in individuals previously vaccinated with a VSV-based vaccine. Using an attenuated variant of ISFV, a vaccine candidate that contains the glycoprotein genes from EEEV FL93–939 strain was generated (Nasar et al. 2017). All CD-1 mice inoculated intramuscularly with a single dose of 10⁸ pfu of ISFV/EEEV developed neutralizing antibodies by day 14. When the immunized mice were challenged intraperitoneally with 10⁵ pfu of virulent EEEV, all ISFV-EEEV animals survived while 9 of 11 control animals succumbed. This preliminary work shows this platform provides an additional candidate option that may be effective with only a single dose. However, additional challenge routes need to be considered with this novel vector.

Subunit Vaccines

Using specific viral proteins that are known to be immunogenic is an approach that has been widely used for a number of reasons including safety and ease of manufacturing. Subunit vaccine products contain no viral nucleic acid nor infectious particles; therefore, they have a clear safety advantage over options that may result in infection even if attenuated or incompletely inactivated. Peptide or protein production is also easily scalable broadening the manufacturing capacity. While the exposed E2 protein might be a natural selection for a subunit vaccine, a number of studies have shown that the E1 protein of alphaviruses, while less immunogenic than the E2 protein, is highly conserved and may function in a pan-alphavirus vaccine approach. Additionally, alphavirus E1 non-neutralizing antibodies have been shown to be protective in animal infection models (Schmaljohn et al. 1982).

One unique approach to a subunit vaccine utilized a combination of cationic lipids, target antigen, and carrier nucleic acids to deliver the protein of interest (Phillips et al. 2014). The ectodomain of the western equine encephalitis virus (WEEV) E1 protein was incorporated into liposome-antigen-nucleic acid complexes (LANACs) and its ability to cross-protect against a heterologous alphavirus was evaluated. These E1-ecto LANAC complexes were used as immunogens in a CD-1 mouse model of EEEV infection. Mice given a subcutaneous prime/boost vaccination regimen (2 wk apart) developed no neutralizing antibodies but were protected from disease when challenged with virulent EEEV strain (FL93–939). Challenge, by either the intranasal or subcutaneous routes, of vaccinated animals 9 wk post boost resulted in a significant decrease in mortality compared with sham-vaccinated controls (9o% vs 0% survival). Curiously, mice challenged only 2 wk after boost had similar mortality rates in both vaccinated and control groups. These data suggest that while immunity can be long-lasting, it may also take some time to develop, which might limit this approach in a public health emergency situation. Additionally, this work demonstrates that cross-protection among alphaviruses can be achieved and should be explored more fully.

Nucleic Acid Vaccines

Nucleic acid vaccines constitute one of the newest categories of vaccine approaches for many viruses, including arboviruses. There are a number of advantages associated with nucleic acid vaccine platforms that make them attractive including ease of production, the safety associated with these products, the limited need for cold chain storage (for DNA), and potentially, a much longer shelf life. DNA-based vaccine products have also been shown to develop both humoral and cell-mediated immunity (Gurunathan et al. 2000). Because the pathway for protection is not characterized for most immunization processes, having multiple immune systems stimulated is a significant advantage. For EEE, both replicon vaccines and DNA plasmid-launched products have been generated and are described here.

One of the EEEV nucleic acid vaccines developed utilized a virus-like replicon particle (VRP) system where the glycoproteins are generated from replicating but incomplete genome particles expressing the non-structural proteins and the glycoprotein (GP) cassette (Reed et al. 2014). When EEEV-GP VRP’s were used to immunize BALB/c mice, all animals generated IgG and neutralizing antibodies by day 28 that increased after a booster dose. Additionally, these mice all survived aerosol challenge with EEEV while control mice universally succumbed. Of particular note, these studies found that the mice were protected even when challenged 6–12 mo after the initial vaccination. Further work on a second animal model, cynomolgus macaques, showed that 11 of 12 of these NHP’s were protected by an aerosol EEEV challenge delivered 60 d post-prime inoculation. However, some clinical signs were present in these animals including fever and an increase in granulocyte count suggesting incomplete protection. Further characterization of this method of protection will be critical for determining the potential for further development.

An alternative to using RNA replicating particles involves the expression of the EEEV glycoproteins via a eukaryotic expression vector intramuscularly injected using an electroporation syringe (Dupuy et al. 2018). Previously, a number of DNA vaccines were found to be poorly immunogenic necessitating the use of adjuvanted large doses and repeated booster doses (Ferraro et al. 2011); this novel delivery approach was postulated to provide a means to alleviate some of these challenges (Sällberg et al. 2015).

A DNA plasmid with the EEEV glycoproteins was used to inoculate BALB/c mice using a prime/2-boost regimen (Dupuy et al. 2018). Mean ELISA antibody titers were found to be significantly greater than background even after a single inoculation but continued to increase after each booster dose. Interestingly, NT antibody titers increased after the first booster dose but were not significantly above background upon initial vaccination nor did they rise further upon a second booster shot. When immunized mice were challenged by aerosol with a virulent virus, no clinical signs developed, and all the animals survived. A comparator challenge using the IND EEEV vaccine resulted in 60% mortality. A similar immunogenicity profile was reported when immunizing New Zealand white rabbits; however, challenge studies in this species or other animal models have yet to be described with this DNA vaccine candidate.

Prospects for Additional Development

Because EEEV cases or outbreaks have been appearing to increase in magnitude and geographic distribution over the past decade, there has been renewed interest in the development of a vaccine to protect humans against this pathogen. Additionally, because EEEV can be infectious by aerosol and the virus is considered a potential biothreat weapon, the US military has also advanced the development of an EEEV vaccine, particularly as part of a trivalent vaccine against all three encephalitic alphaviruses. While a single target EEEV vaccine is unlikely to be sufficiently attractive to merit advanced development, those approaches that allow for multi-alphavirus targeting could be strongly considered.

Several of the approaches described in this review do address the single target issue and have been developed against the related alphaviruses Venezuelan and western equine encephalitis. The SINV chimeric approach showed that a SINV combined with the structural proteins of Madariaga virus (MADV) would protect against EEEV challenge (Roy et al. 2013). The DNA, VRP, VLP, ISFV, and EILV vaccine approaches all use a mixture of vaccines against eastern, Venezuelan, and/or western equine encephalitis virus components to attempts to protect against homologous or heterologous virus challenges (Reed et al. 2014, Nasar et al. 2017, Dupuy et al. 2018, Erasmus et al. 2018, Ko et al. 2019). These products have varying degrees of cross-protection depending upon the approach, challenge strain, and route. However, all of them have demonstrated sufficient promise to warrant further evaluation. One product has even been developed that contains a combination of antigens from the three encephalitic alphaviruses within a single expression vector, the MVA candidate (Hu et al. 2018). This single vector containing multivalent antigens provides an additional advantage of not requiring optimization of a mixture of vaccine products. All of these multivalent approaches are also attractive as they provide broad protection against multiple viruses rather than necessitating the development of vaccines for each individual target.

Unfortunately, advanced vaccine development to bring a product to market is contingent upon many factors other than science. For EEEV, marketing a vaccine is particularly challenging given that outbreaks are generally small, can be infrequent, and are geographically difficult to predict. These factors can hinder motivation for further vaccine development due to financial constraints. To develop and license a new vaccine product may take from 5 to 18 yr and can cost as much as $500 million (Plotkin et al. 2017) making the development of vaccines for orphan viruses impractical. However, given that EEEV is a potential biowarfare agent and the cost of long-term care for individuals surviving the neurological disease is extensive (Wolfe et al. 2014), additional considerations for development exist that are not always present for most pathogens.

One advantage in evaluating many of the EEE vaccine products that have been produced is that direct comparisons are able to be made. The same virus strains, animal models, and routes of delivery are used for many of these choices making the studies comparable. Additionally, many studies include the same or similar evaluations of effectiveness including determining PRNT titers at frequent time points (although not always the same time points) and level of viremia upon challenge. Having the scientific community agree upon certain parameters for product evaluation greatly facilitates direct comparisons of the promise of each option. An additional positive element of many of these candidate evaluations is that they were evaluated for protection by more than one challenge route. Because EEEV is primarily transmitted by mosquito bites yet could be infectious by an aerosol release, having data on protection for both scenarios is critical for further development.

As with other vector-borne viruses, such as chikungunya virus, the science used in all of these approaches described for an EEEV vaccine is sound and many have been examined for potential use against other alphaviruses (Powers 2018). The primary barriers to any eventual licensed product remain financial, political, and related to the feasibility of completing the requirements for licensure.