vaccines || dengue vaccines

44SECTION THREE: Vaccines in development and new vaccine strategies

Dengue vaccines Scott B. Halstead * Stephen J. Thomas

Since World War II, the dengue viruses (DENVs) have spread throughout tropical and many subtropical areas, achieving pandemic status and causing tens of millions of overt infec-tions annually. Of these millions, more than 500,000 persons are hospitalized with severe illnesses, including dengue hem-orrhagic fever (DHF) and dengue shock syndrome (DSS), and 0.1% to 5% die. 1–3 Dengue exacts a high socioeconomic toll on inhabitants of more than 100 tropical countries ( Figure 44-1 ). In addition, tourists become infected. In 2007, 141 million persons traveled to dengue-affected areas. 4 Dengue illnesses exact a severe impact on quality of life. 5 In Asia and the Americas, the economic burden of dengue on the poor has been estimated at 1,289 disability-adjusted life years per mil-lion population, similar to the burden imposed by the child-hood and tropical disease clusters and tuberculosis. 6–9 An eight-country survey found costs for both out-patient and in-patient dengue illnesses to be substantial, ranging from 8 to 56 days of gross domestic product (GDP) per capita, costing the studied countries an estimated $587 million (in 2005 dol-lars) per year. 10 The global burden of dengue is both significant and underestimated. 11

Dengue viruses and dengue disease

There are four dengue viruses (DENV-1 through -4), members of the Flavivirus group of the virus family Flaviviridae. The flavivirus genome consists of approximately 11,000 base pairs, which translate into three structural and seven non-structural proteins (see “Chimeric virus vaccines”, later). The four distinct dengue viruses all evolved from a com-mon sylvatic ancestor, with separate introductions into the urban cycle of transmission—human to Aedes mosquitoes to human. 12 The basic virology of the four DENVs is similar to

that of yellow fever (see Chapter 38) and Japanese encephali-tis (see Chapter 19 ). Infection of humans follows the bite of an infected Aedes mosquito (primarily Aedes aegypti, Aedes albopictus, or Aedes polynesiensis ). Infection with DENV-1 (five genotypes), DENV-2 (six genotypes), DENV-3 (five geno-types), or DENV-4 (four genotypes) may be asymptomatic, or it may result in a mild febrile illness, dengue fever (DF), or severe dengue including DHF/DSS. 13 Dengue fever is a self-limited febrile illness, with an average incubation period of 5 days, characterized by acute onset of fever, headache, myal-gia, and arthralgia. The continuum from DF to severe den-gue is differentiated physiologically by the degree of vascular permeability, which occurs around the time of defervescence, when virus-infected cells are being immunologically elimi-nated. Hemorrhage, mild or severe, may be a component of either DF or DHF. Cytokines and chemokines generated by immune elimination and capable of affecting endothelial cell and platelet integrity circulate in blood at levels proportionate to disease severity. 14

Dengue illnesses are documented either by detection of the etiologic agent or by documenting a specific antibody response after infection. Determination of etiology during the acute phase centers on isolating virus in any of several host systems or detecting circulating dengue RNA. An alternative method, recently commercialized, is the detection of dengue nonstructural protein 1 (NS1), which circulates in the blood in the acute and early convalescent phases. 15–17 Many tests, some commercialized, can be used to measure antibody responses to dengue infection. The most common is the detection of enzyme-linked immunosorbent assay (ELISA) IgM antibodies in late acute or early convalescent sera. 18 The US Food and Drug Administration (FDA) allowed marketing of the first test to detect dengue (IgM Capture ELISA) in April 2011. 19 It may also be important to distinguish primary dengue infections from infections that occur in persons immune to another den-gue virus or to other flaviviruses (secondary infections). This is done by detecting IgG ELISA antibodies in acute-phase sera, and by comparing the ratio of IgM to IgG dengue antibodies in acute or early-convalescent phase sera. 20

Classical DHF/DSS accompanies first DENV infections in 5 to 11 month old infants born to dengue-immune mothers, 21 and during second (occasionally third) heterotypic DENV infections in older persons. 22 These two epidemiologic settings have in common a single immune factor—IgG1 dengue anti-bodies. In vitro and in vivo studies have shown that dengue antibodies at subneutralizing concentrations enhance DENV infections in Fc-receptor-bearing cells, a process referred to as antibody-dependent enhancement. 23 , 24 Enhanced vire-mia titers correlate with severe second DENV infections in

* Disclaimer: The opinions or assertions contained herein are the private views of the author (S.J.T.) and are not to be construed as reflecting the official views of the United States Army or the United States Department of Defense. Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adhered to the principals stated in the Guide for the Care and Use of Laboratory Animals, NRC Publication, 1996 edition. One author (S.J.T.), as an employee of the United States Army, has been assigned to work on dengue vaccine co-development efforts with numerous commercial entities, some mentioned in this article. The author discloses these relationships not because there is a conflict of interest but for transparency.

1043Dengue vaccines 44

humans. 25 Cellular immune responses after DENV infection are postulated to contribute to observed immune and clinical pathology. 26–29

The DHF/DSS response is rare, seen in only 2% to 4% of second DENV infections. 30 Although supportive intensive care and careful fluid management can be life saving, specific anti-viral medications are not available. 31 The single preventive measure, comprehensive mosquito control, has conspicuously failed. 32

Developing dengue vaccines

The numerous challenges that face the dengue vaccinologist have made the goal of developing a safe and efficacious den-gue vaccine elusive for more than four decades. 33 Each DENV type is capable of causing disease and death, and many den-gue-endemic areas have concomitant cocirculation of mul-tiple DENV types, mandating development of a tetravalent (DENV-1 + 2 + 3 + 4) dengue vaccine. There is no validated animal model of dengue disease and no validated human chal-lenge model, making vaccine candidate down-selection dif-ficult and often requiring numerous, small, phase 1 human studies.

After natural infection with DENV-1, a few months of cross-protection against infection with DENV-2 occur, after which heterotypic infection often results in overt illness. 34 With the risk of severe dengue increased during a second DENV heterotypic infection, a single-dose tetravalent vaccine would be ideal for a dengue vaccine. 23 , 35 Because classical DHF has occurred in DENV-1–immune adults infected with DENV 2 after an interval of 20 years, it must be assumed that immu-nologically primed persons are at risk for enhanced disease for a lifetime. 36 The immune profile driving immunoprotective or

immunopathogenic responses is complex and incompletely understood. The biologic assays used to measure primary immunogenicity readouts (eg, neutralizing antibody) are highly variable, lack robustness, and are cross-reactive among the DENVs, making it difficult to know what responses are type-specific or heterotypic. 37

The success of candidate dengue vaccines may well hinge on neutralizing antibody. The convalescent immune response after infection with a single DENV type results in decades, per-haps life-long measurable neutralizing antibody, resulting in solid protection against challenge with the same virus type. 34 , 38 Numerous studies suggest that neutralizing antibody elic-ited by active vaccination prevents viral replication (thus pro-vides protection) after challenge in nonhuman primates. 39–42 Similarly, volunteers who developed neutralizing antibody after vaccination with candidate live attenuated virus dengue vaccines were protected after challenge with underattenuated viruses. 43 Because of these features, neutralizing antibody has become the primary immunogenicity readout for the assess-ment of dengue vaccine candidates, and the plaque reduction neutralization test, or microneutralization platform variant, has become the means of measurement. 44 , 45 Vaccine develop-ers have also committed resources to measuring B- and T-cell memory and other cellular immune responses after vaccina-tion, but it is unlikely that these readouts (eg, B-cell enzyme-linked immunosorbent spot [ELISPOT], intracellular cytokine staining, cytometric bead array, interferon-gamma ELISA) will be the primary immunogenicity readout supporting vaccine licensure. 46–49

Recent reviews of dengue vaccine development admirably summarize the published literature and patent filings. 50–52 The World Health Organization has published results of working group discussions of technical specifications for the manufac-ture and evaluation of dengue vaccines. 53

10,000–100,000 DF/DHFcases during a single year

A decade of Dengue:1995–2004

100–10,000 DF/DHFcases during a single year

<100 DF/DHFcases during a single yearor no report

>100,000 DF/DHFcases during a single year

Figure 44-1 Dengue fever and dengue hemorrhagic fever shown by country and number of cases reported to the World Health Organization or its regional offices, from 1995 to 2004. Dengue infections occur widely in West Africa and in coastal East Africa, but few cases are reported. (From the World Health Organization DengueNet [ www.who.int/csr/disease/dengue/denguenet ] and the Pan American Health Organization [ www.paho.org/english/ad/dpc/cd/dengue.htm ].)

1044 SECTION THREE • Vaccines in development and new vaccine strategies

Vaccines in development

The development of dengue vaccines began in 1929 with unsuc-cessful efforts to produce an immunogenic, inactivated virus vaccine using phenol, formalin, or bile.54,55 During World War II, live attenuated virus (LAV) vaccines were invented in two labo-ratories independently by passaging DENV 1 serially in suck-ling mouse brain.56,57 DENV 2 was similarly attenuated by serial mouse brain passage.58 These early live virus approaches were abandoned because of concerns about the safety of inoculating mouse tissues into humans.59 Since the 1970s, numerous new dengue vaccine constructs have entered preclinical and clinical development. Currently, six different vaccine approaches have been tested in human clinical trials, and a single candidate is now in phase 3 clinical testing. Feasibly, a dengue vaccine could be licensed in endemic countries within 5 years.

Live attenuated virus vaccines

Selection of viral mutants

Historically, flaviviral LAV vaccines have proved to be safe and durably efficacious. Examples include the 17D vaccine for yel-low fever60,61 and the SA 14-14-2 vaccine for Japanese encepha-litis.62,63 Live attenuated viral vaccines replicate in the recipient, presenting all viral antigens and eliciting both antibody and T-cell responses, resembling those seen after natural infection. A tetravalent vaccine approach is designed to induce primary-type immune responses to all four dengue viruses simultaneously, a phenomenon demonstrated to be protective in susceptible rhe-sus monkeys.39,64 In fact, most adults in Southeast Asia during their lifetimes experience DENV infections sequentially, and they have circulating neutralizing antibodies to all four DENV types with demonstrable solid immunity but rarely severe den-gue disease. The reason for this is not known but probably related to individual genetic factors and the protection afforded by heterotypic dengue neutralizing antibodies.

The first major effort at live attenuated dengue vaccine devel-opment was made at Mahidol University in Bangkok; investi-gators used the classical method of serial passage of virus in a nonhuman host. Virus strains DENV-1 16007, DENV-2 16681, and DENV-4 1036 were passaged 15 times in primary dog kid-ney (PDK) cells at the University of Hawaii, and then trans-ferred to Thailand for further passage, development of candidate

vaccines, and phase 1 testing.65 Extensive studies in small groups of flavivirus-susceptible Thai adult volunteers who were given various PDK-passage-level viruses resulted in strains with acceptable reactogenicity and immunogenicity for DENV-1, -2 and -4 viruses at passage levels PDK13, PDK53, and PDK48, respectively.66–68 DENV-3 viruses failed to replicate in PDK cells and were attenuated by 48 passages in primary African green monkey kidney cells (DENV-3 16562), and three final passages in fetal rhesus lung cells. All 10 US Army soldiers inoculated with DENV-2 PDK 53 developed neutralizing antibodies.69

In subsequent studies, the four monovalent candidates elic-ited neutralizing antibody seroconversions in 3/5, 5/5, 5/5, and 5/5 American volunteers, respectively, after a single dose of 103 to 104 plaque-forming units (pfu).70 Bivalent and trivalent formu-lations using DENV-1, -2 and -4 vaccine candidates elicited bal-anced seroconversions.71 However, when all four serotypes were combined into a tetravalent vaccine, the predominant response was to the DENV-3.70 The vaccine was reformulated using lower viral concentrations of DENV-3 and resulted in more balanced tetravalent seroconversion.72,73 Children vaccinated were followed for up to 8 years without an observed increase in risk for severe disease.74 Despite these promising results, all 10 adult Australian volunteers given tetravalent vaccine had DENV-3-related systemic reactions, including fever, headache, arthralgia/myalgia, eye pain, and rash—essentially, mild DF.75 Similar results occurred in adult Asian volunteers inoculated with a molecularly rederived DENV-3 strain, which was selected to reduce reactogenicity but had the reverse effect.76,77 Further testing and development of the Mahidol live attenuated vaccine has now stopped.

A second tissue-culture-passaged vaccine developed at the Walter Reed Army Institute of Research (WRAIR) was derived by serial passage of all four dengue viruses in PDK cells with final passages in fetal rhesus lung cells (Table 44-1). Ten mon-ovalent candidates were tested in 65 flavivirus-naïve volunteers between 1987 and 1996.78–82 Vaccine candidates (specific virus strains, viral concentrations, and PDK cell passage levels) were evaluated. Candidates DENV-1, 45AZ5 PDK-20; DENV-2, 16803 PDK-50; DENV-3, CH53489 PDK-20; and DENV-4, 341750 PDK-20 with single-dose seroconversion rates of 100%, 67%, 50%, and 63%, respectively, and acceptable reactogenicity were selected for continued development.

In a follow up study 49 flavivirus-naïve adult volunteers were administered the monovalent candidates.83 Neutralizing antibody seroconversion rates for DENV-1, -2, -3, and -4 were 100%, 92%, 46%, and 58%, respectively. A second dose adminis-tered 1 to 3 months later failed to improve immunogenicity—an

Type of vaccine Dengue virus genes (N) Stage of development

Live attenuated virus (traditional) 10 (all) Phase 2 tetravalent (Walter Reed Army Institute of Research and GalaxoSmithKline [GSK])

Live attenuated virus (molecular) 10 Phase 2 monovalent (US National Institutes of Health)Phase 1 tetravalent (US National Institutes of Health)Protects nonhuman primates (US Food and Drug Administration)

Yellow fever chimera Chimera, 2 + 8 (yellow fever virus) Phases 2b to 3 tetravalent (Sanofi Pasteur)

Dengue chimera Chimera, 2 + 8 (DENV-2) Phase 1 tetravalent (Inviragen)

Purified inactivated 3 Protects monkeys (GSK, Walter Reed Army Institute of Research)

Recombinant subunit < 1 Phase 1 DENV 1 (Hawaii Biotechnology, Inc.)

DNA 2+ Protects monkeys (Naval Medical Research Center and Maxygen)

Table 44-1 Development of Dengue Fever Vaccines


observation reminiscent of Sabin's observed period of cross-protection after primary challenge. 34 , 84 There were no serious adverse reactions. The DENV-1 candidate was the most reac-togenic, with a 42% rate of fever ( > 100.4° F) after a single dose, and reactogenicity was uniformly less after dose 2.

The monovalent candidates were combined into a tet-ravalent preparation and administered to 10 volunteers. 83 Seroconversion rates after a single dose of tetravalent vaccine by DENV type were 70%, 60%, 50%, and 30%, respectively. Dose 2 failed to improve antibody response but dose 3 improved sero-conversion rates—100%, 75%, 100%, and 50%, respectively. There were no serious safety concerns.

Tetravalent vaccines made using combinations of high (10 5,6 pfu/dose) or low (10 3.5-4.5 pfu/dose) viral concentration formula-tions of each of the four DENVs were inoculated in 64 flavivirus naïve adult volunteers. 85 Formulations 1 to 15 were administered to 54 volunteers, three to four volunteers per formulation, on days 0 and 28. Formulation 16 was tested in 10 volunteers, five volun-teers inoculated on days 0 and 30, one volunteer on days 0 and 120, and four volunteers on days 0, 30, and 120. Reactogenicity was variable across the formulations and positively correlated with immunogenicity. Similar proportions of volunteers seroconverted to DENV-1, -2, and -3 (69%, 78%, and 69%, respectively), but significantly fewer volunteers seroconverted to DENV-4 (38%). Seven formulations were found to be serologically superior and three of these sufficiently nonreactogenic to justify further testing.

The WRAIR entered a codevelopment partnership with GlaxoSmithKline (GSK) Biologicals to further develop tetrava-lent vaccine candidates. Additional monovalent components (DENV-1, 45AZ5 PDK-27 and DENV-4, 341750 PDK-6) were used to make a 17th formulation to reduce reactogenicity and improve the balance of the immune response. Three formula-tions (13, 14, and 17), consisting of combinations of 10 3-5 pfu/dose of each DENV type, were studied in 71 adults. 86 Earlier studies demonstrated the need to extend the immunization schedule and the requirement for multiple dosing to achieve optimal antibody responses (balance and potency). Thus, vol-unteers in this study received two doses at 0 and 180 days. Formulation 17 was selected for further clinical evaluation.

Because DHF/DSS is regularly seen in infants 2 years of age and older, tetravalent vaccines must be given early in life. It was reasoned that vaccines somewhat reactogenic in adults might be acceptable in children, because in hospital-based studies in Thailand most primary DENV infections in young children are clinically silent. 25 , 87 Thus, formulation 17 was taken forward into a small safety and immunogenicity study in flavivirus-naïve Thai schoolchildren ( N = 7), 6 to 9 years of age. Two doses, given 6 months apart, were well tolerated and resulted in 100% of the cohort developing tetravalent neutralizing antibodies. 88 A subsequent study of the same formulation was conducted in flavivirus-naïve Thai toddlers ( N = 34 vaccine recipients) 12 to 15 months of age. Vaccination was well tolerated and, although variable neutralizing antibody responses were observed across cohorts, sufficiently immunogenic to pursue continued devel-opment of the candidate. 89

New tetravalent dengue vaccine (TDEN) lots were produced to try to reduce the theoretical risk of transmissible spongi-form encephalopathy arising from use of animal products of incompletely documented origin in the vaccine seeds or dur-ing decades of cell culture. TDEN lots were tested in phase 2 studies. Studies in the United States and Thailand were con-ducted in adults, and in Puerto Rico across a broad age range (12 months to 50 years). Screening was not performed, and flavivirus-primed and naïve subjects were enrolled. There were no overt safety signals in over 300 vaccine recipients, and rates of seroconversion were moderate to high. Dengue-primed vol-unteers experienced no increased reactogenicity but did experi-ence more robust and broader neutralizing antibody responses after a single dose. 90

The future of the WRAIR/GSK TDEN vaccine in its current form is in question because of manufacturing complexities and the desire by both organizations to shorten the time to protec-tion by altering the dosage schedule, or to explore other vaccine approaches.

Directed mutagenesis

National Institutes of Health vaccine The National Institutes of Health (NIH) introduced a new era of dengue vaccine research when they constructed viable cDNA clones of DENV-4 (WRAIR 814669, Dominica, 1981) (see Table 44-1 ). 91 , 92 A DENV-4 mutant virus (rDENV-4 ∆ 30) with a 30-nucleotide deletion in the 3′ untranslated region (10478-10507) produced lower viremia levels than parental DENV-4, but the neutralizing antibody responses were comparable to those of parental DENV-4 when inoculated in susceptible rhe-sus monkeys. 93 Phase 1 trials in flavivirus-naïve volunteers showed low-level viremia; frequent mild, largely asymptom-atic rashes; infrequent elevations in alanine transferase blood levels; occasional depression of absolute neutrophil count; and high levels of neutralizing antibodies at dosages of 10 1 to 10 5 pfu/0.5 mL. 94 , 95 A. aegypti mosquitoes fed on vaccinees did not become infected. 96 By screening mutants, a DENV-4 with paired amino acid substitutions at positions 200 and 201 of NS5 was obtained. 97 This virus showed reduced replication in severe combined immunodeficiency mouse, human hepatoma, and rhesus monkey models. 95 , 98 In flavivirus-naïve humans, inoculation of this virus abrogated hepatotoxicity and viremia but retained full immunogenicity. 99 Introduction of ∆ 30 into DENV-1 resulted in a successful vaccine virus. 100 Successful human vaccines for DENV-2 were obtained only by use of chi-meras with rDENV-4 ∆ 30. 98 , 101 , 102 A DENV-3 human vaccine candidate was obtained by substituting the entire 3′ UTR with that of rDENV-4 ∆ 30. 103

Tetravalent formulations containing these viruses along with rDENV-1 ∆ 30 and rDENV-4 ∆ 30 have been tested in monkeys and shown to be attenuated, and they elicit balanced antibody responses. 41 In total, eight monovalent vaccine candidates were tested in 15 phase 1 clinical trials to identify optimal candidates for tetravalent preparations. A subcutaneous injection of 10 3 pfu induced neutralizing antibody seroconversions to the vac-cine parent virus in 80% to 100% of vaccine recipients. 104

Tetravalent candidates (admixtures, TetraVax-DV) were prepared using the safest and most immunogenic monova-lent candidates. A phase 1, double-blinded, randomized, placebo-controlled trial of three TetraVax-DV admixtures was recently completed in flavivirus-naïve US adults. Admixture 1 is composed of 10 3 pfu of rDEN-1 ∆ 30, 10 3 pfu of rDEN-2/4 ∆ 30[ME], 10 3 pfu of rDEN-3-3′D4 ∆ 30, and 10 3 pfu of rDEN-4 ∆ 30. Admixture 2 includes 10 3 pfu of rDEN-1 ∆ 30, 10 3 pfu of rDEN-2/4 ∆ 30[ME], 10 3 pfu of rDEN-3-3′D4 ∆ 30, and 10 3 pfu of rDEN-4 ∆ 30-200,201. Admixture 3 includes 10 3 pfu of rDEN-1 ∆ 30, 10 3 pfu of rDEN-2/4 ∆ 30[ME], 10 3 pfu of rDEN-3 ∆ 30/31-7164, and 10 3 pfu of rDEN-4 ∆ 30. 104–106

The NIH has licensed vaccine candidates to Butantan Institute, Sao Paolo, Brazil; Biological E Ltd, Hyderabad, India; Panacea Biotech, Ltd, New Delhi, India; and Vabiotech, Hanoi, Vietnam.

US Food and Drug Administration vaccine Investigators at the Center for Biologics Evaluation and Research at the FDA created molecularly attenuated DENV-1 (DENV-1 WP parent) and DENV-2 (DENV-2 NGC parent) strains by replacing a portion of the wild-type DENV termi-nal 3′ stem-and-loop structure with that of West Nile virus (see Table 44-1 ). The chosen chimeric virus grew normally in mammalian LLC-MK2 cells, but its growth was severely


restricted in C6/36 insect cells; it was designated mutant F (or mutF).107,108 When inoculated into susceptible rhesus monkeys, viremia was greatly reduced compared with paren-tal virus, but the neutralizing antibody responses were good. Monkeys challenged 17 months after a single dose of vaccine were protected from viremia.108 This molecularly attenuated virus provides an alternative approach to constructing a tetra-valent dengue vaccine.

Chimeric virus vaccines

The flavivirus genome is a single-stranded, positive-sense RNA molecule of nearly 11 kilobases containing a single open read-ing frame. The RNA is translated into a polyprotein that is processed into at least 10 gene products: the three structural proteins (nucleocapsid or core [C], premembrane [prM], and envelope [E]) and seven nonstructural (NS) proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5). The untranslated regions of the genome at the 5′ and 3′ ends are crucial for pro-tein translation and minus-strand transcription.109,110 The E protein participates in virus and host-cell membrane fusion and subsequent infection. As such, DENV-type specific neutraliz-ing antibodies directed against the DENV E and possibly other structural proteins should be a major protective component of the convalescent immune response.111

Chimera of dengue and yellow feverInsertion of dengue preM and E genes into the cDNA back-bone of yellow fever (YF) 17D was pioneered at the St. Louis University Health Sciences Center, further developed by Acambis, Inc., and licensed for manufacture to Sanofi Pasteur (Figure 44-2, and see Table 44-1).70,112–114 Studies to date sug-gest that chimerization of dengue and West Nile viruses pro-vides attenuation.113,115–118 Furthermore, these vaccines benefit from the high fidelity of YF 17D polymerase.119 However, for

construction of YF 17D–based vaccine viruses against encepha-litis viruses (eg, Japanese [JE], tick-borne, and St. Louis enceph-alitis viruses), additional attenuation of the prM-E genes of the vaccine viruses was necessary.113,120–123 Vero cells serve as the substrate for vaccine virus production.

Chimeras of dengue and yellow fever viruses have been developed using the PrM-E genes from DENV-1, PUO-359; DENV-2, PUO-218; DENV-3, PaH881; and DENV-4, 1228 viruses.124 Neurovirulence in mice and monkeys was reduced compared with the YF 17D vaccine virus.125,126 Viremia in rhe-sus macaques is similar to that of YF 17D virus and greatly reduced compared with wild-type DENVs. In rhesus mon-keys given a tetravalent formulation comprising 3 logs of the DENV-2 chimera and 5 logs of the DENV-1, -3, and -4 chime-ras, the seroconversion rate was 100%,42,124 and monkeys had no measurable viremia after challenge with all four wild-type DENVs.124 Chimeric viruses replicate poorly in potential vector mosquitoes.127,128

Phase 1 data for the YF 17D–Japanese encephalitis virus chimera (ChimeriVax-JE) provided initial clinical proof of principle for the yellow fever chimeric platform.129 Safety and immunogenicity of a monovalent DENV-2 candidate, ChimeriVax-DENV-2, was established in a phase 1 study.130 A single subcutaneous dose of the investigational vaccine (103 or 105 pfu) was tested in volunteers who were yellow fever anti-body naïve or who had previously received yellow fever vaccine. Most adverse events were similar to those seen with licensed yellow fever vaccine, and of mild to moderate intensity with no serious side effects. Of yellow fever–naïve subjects inoculated with 105 or 103 pfu of ChimeriVax-DENV-2, 100% and 92.3%, respectively, seroconverted to wild-type DENV-2 (strain 16681). Observed viremia was transient and low in titer, not exceed-ing the levels from yellow fever vaccination. Prior immunity to yellow fever did not interfere with DENV-2 neutralizing anti-body responses but induced a long-lasting and cross-neutralizing antibody response to all four DENV types.130

C

C

prM

prM

E

E

Non-structural genes

prM E

Non-structural genes

Yellow fever V 17D cDNA

Exchange with genes of wt dengue 1--4

4 chimeric cDNAs

Individuallytranscriptedto RNA

Virus grownin Vero cells

RNA transfection

Four individual chimericDengue viruses (CYD1–4)

1 PUO-359/TVP-1140

2 PUO-218

3 PaH881/88

4 1228 (TVP-980)

1 2

3 4

Figure 44-2 Construction of ChimeriVax-based vaccines. Chimeric flavivirus vaccines are constructed by replacing the genes coding for prM and E proteins from yellow fever virus (YFV) 17D 204 vaccine with those of heterologous flaviviruses (Dengue, West Nile, or Japanese encephalitis). After DNA cloning, RNA is transcribed and transfected into Vero cells to obtain chimeric viruses possessing the YFV 17D replication machinery and the external coat of the relevant heterologous flavivirus.


Different formulations of chimeric tetravalent candi-dates produced significantly different neutralizing antibody responses in rhesus monkeys, suggesting an interference phe-nomenon. 124 This phenomenon was bypassed in susceptible monkeys by administering bivalent vaccines in two separate sites. 131 Individual animals with preexisting flavivirus anti-bodies, either in or outside the DENV group, respond to a single administration of a tetravalent chimeric vaccine with a broad tetravalent neutralizing antibody response. 131 The tet-ravalent formulation ChimeriVax-DENV was given to healthy adult volunteers randomized into two groups. 132 Group 1 received three tetravalent dengue vaccines (TDVs) injections at months 0, 4, and 12 to 15; group 2 received saline placebo at month 0 and then two TDV injections at months 4 and 12 to 15. Adverse events were recorded, and biological parameters and viremia levels were measured. Neutralizing antibodies against World Health Organization (WHO) reference strains were measured before and after vaccinations. Among 33 par-ticipants in each group, low-level viremia was detected, mainly in the DENV-4 group. The neutralizing antibody responses to each of the four DENV types increased with successive vaccine doses. All participants seroconverted against all four serotypes after receiving three doses at 0, 4, and 12 to 15 months, and almost all seroconverted after two doses given between 8 and 11 months . 132

Volunteers of various ages (2 to 45 years), genetic back-grounds (United States, Mexico, Australia, Philippines), and flavivirus priming status have demonstrated no overt safety sig-nals, with safety profiles comparable to those of the licensed control vaccines. A US adult cohort had a 100% tetravalent seroconversion rate after three doses of DENV-1, -2, -3, or -4, with geometric mean titers of 67, 538, 122, and 154, respec-tively. 132 A flavivirus-naïve cohort from Mexico (2 to 45 years old) experienced 77% to 92% seroconversion rates after three doses, and the rates were 88% to 100% in the 2 to 5 year old subgroup. 133 Cellular immune responses measured during clini-cal development were consistent with T-helper cell type 1 (Th1) responses, dominated by interferon-gamma for both CD4 and CD8 responses. 46 , 132 , 134 Regarding a favorable or unfavorable role of T cells in vaccinated subjects, it has been observed that there was no cross-reactivity between DENV NS3– specific and YF NS3–specific CD8 + T cells. 46 , 47

A randomized, controlled, multicenter, phase 2 study in Mexico City with children aged 2 to 5, 6 to 11, and 12 to 17 years (36 children per age group) and with 18 adults 45 years old or younger received three injections of TDV at months 0, 3.5, and 12 (TDV-TDV-TDV), or 1 injection of yellow fever vac-cine (YF) at month 0 and two injections of TDV at months 3.5 and 12 (YF-TDV-TDV). 133 Plaque reduction neutralization test antibody titers against the TDV parental viruses were measured 28 days after vaccination. Seropositivity was defined as antibody titers greater than or equal to 1:10. No vaccine-related serious adverse events, other significant clinical adverse events, or clinically significant trends in biological safety were observed. Reactogenicity did not increase with successive TDV injections. Mild-to-moderate injection site pain, headache, myalgia, and malaise were only slightly more common than in controls (14% to 40% after each vaccination). After three TDV vaccinations of susceptible persons, the seropositivity rate against each dengue serotype was in the range 77% to 92%, compared with 85% to 94% after completion of the YF-TDV-TDV regimen. Of the 2 to 11 year old participants, 95% were seropositive against three or more serotypes after three vaccinations. 133

Sanofi Pasteur is completing a proof-of-concept efficacy trial and a large-scale safety trial of three doses (at 0, 6, and 12 months) of their tetravalent dengue–yellow fever vac-cine compared with placebo in a cohort of 4,000 flavivirus- susceptible and partially immune children (4 to 11 years) in Ratchaburi, Thailand. 135 A phase 3 lot-consistency trial at

the same dosage regimen was recently initiated in 715 largely flavivirus-susceptible Australian adults. 136 A second phase 3 trial is ongoing to assess safety and immunogenicity in Malaysian children (aged 2 to 11 years). 137 Two additional phase 3 trials are being initiated in dengue-endemic areas of Latin America and Asia. 138–140 As documented in a series of publications, Sanofi Pasteur hopes to complete clinical trials required for registration and licensing of their vaccine in dengue-endemic areas as early as 2015. 47 , 138 , 141–143 As of September 2011, more than 11,000 volunteers had received at least one dose of a monovalent or tet-ravalent ChimeriVax-DEN (J. Lang, personal communication, 20 September 2011).

Dengue-dengue chimera The Centers for Disease Control and Prevention (CDC) devel-oped a tetravalent chimeric dengue vaccine by introducing DENV-1, -3, and -4 prM and E genes into cDNA derived from the successfully attenuated DENV-2 component of the Mahidol University–Sanofi Pasteur LAV vaccine (DENV-2, 16681 PDK-53) (see Table 44-1 ). 69 , 144 , 145 Attenuating mutations of the DENV-2 vaccine are located in the nonstructural genes and reportedly are stable. 146–149 DENV-1–DENV-2 chimeras have been produced using DENV-1 PDK-13 and parental DENV-1 16007 structural genes. 147 DENV-3–DENV-2 and DENV-4–DENV-2 chimeras using parental virus structural genes are stable in cell culture, and when all constructs were inoculated in cynomolgus monkeys, monomeric and tetravalent formula-tions resulted in neutralizing antibody responses, low viremia titers, but some evidence of interference. 146 , 150

Dengue-dengue chimeras were formulated as a DENV-1/DENV-2, DENV-2, DENV-3/DENV-2, and DENV-4/DENV-2 tetravalent vaccine candidate and licensed to Inviragen, Inc. The vaccine candidate (DENVax) has undergone preclinical testing in mice and nonhuman primates, comparing intrader-mal and subcutaneous deliveries. The intradermal delivery of DENVax (10 5 pfu per DENV type) produced superior immune responses in cynamolgous macaques; viremia was not detected after challenge with 10 5 pfu of DENV-1 West Pacific or DENV-2 New Guinea C viruses (J. Osorio, personal communication, 2010). Phase 1 clinical trials in flavivirus-susceptible adult vol-unteers are being conducted at St. Louis University School of Medicine 151 and in Medellin, Colombia. 152

Exploring the safety of replicating dengue vaccines

After the administration of live virus dengue vaccines, two safety issues should be considered: (1) preexisting dengue or nondengue flavivirus antibodies in the vaccinated person may lead to increased vaccine virus reactogenicity, and (2) in the event of primary or secondary vaccine failure or waning neutralizing antibody over time, severe disease (ie, enhanced infection and disease) may accompany wild-type DENV infec-tions. Absence of validated animal models of dengue disease and human challenge models has made it necessary to rely on clinical studies to address these questions. The small datasets available at this time do not test the occurrence of these hypo-thetical outcomes.

Primary infections with some strains of wild-type DENV-4 and DENV-2 viruses do not cause overt disease in children and cause them only occasionally in adults. By contrast, some of the same strains, particularly of Asian origin, may produce severe disease during a secondary DENV infection. 25 , 153 , 154 The implication for live virus vaccines is that they too may appear to be attenuated when inoculated in dengue-naïve volun-teers but “virulent” in persons partially immune to dengue. To date, neither live attenuated nor live chimeric dengue vaccines when inoculated into dengue-immune children or adults have resulted in enhanced disease caused by vaccine virus. 72 , 73 , 138

1048

SECTION THREE • Vaccines in development and new vaccine strategies

Among more than 140 persons who received tetravalent dengue vaccines of varying immunogenicity, there is no indica-tion of more severe disease after documented wild-type DENV infections remote from vaccination (S. J. Thomas, personal communication, 2010).74 Nor has increased disease severity been observed in dengue-vaccinated persons immune to yel-low fever,155 in volunteers receiving closely or remotely spaced sequential monovalent live dengue virus vaccines (W. Sun, per-sonal communication, 2010),34,156 in clinical trials in which two doses of tetravalent vaccine were given 6 months apart,85 or in dengue- or JE-immune volunteers given serially passaged, mono-typic or tetravalent DENV vaccines (S. J. Thomas, personal communication, 2010).78,157 Under natural conditions, more than 95% of secondary dengue virus infections are mild or inap-parent.30,158 To adequately assess this risk and the risk of incom-plete immunization and waning antibody titers, dengue vaccine clinical development plans must include flavivirus- primed and flavivirus-naïve volunteers, and sufficiently long-term follow-up to make statistically powered conclusions regarding the safety of dengue vaccination in flavivirus-endemic areas.159

Additional safety concerns with live virus dengue vaccines include the presence of cell culture–derived adventitious agents in the final vaccine product, community spread of vaccine virus via vector mosquitoes, vaccine virus neurovirulence, the effects of vaccine administration in immunocompromised hosts, and recombination events. The FDA has published guidance on characterization and qualification of cell substrates and other biological materials used in the production of viral vaccines, with a description of how to address potential adventitious agents.53,160

Risk assessments of vector transmission by vaccine recip-ients have also been performed by numerous dengue vac-cine developers. Results of published studies indicate a very low likelihood that a vaccinee could transmit vaccine-derived dengue viruses to a mosquito.117,127,128,161,162 Because of the neurovirulence of dengue viruses passaged in mouse brain, there is a long record of DENV neurovirulence assessment of DENV vaccine strains in mouse and nonhuman primate models.56 The results have demonstrated that neurovirulence is absent in parental as well as cell culture–passaged DENV strains.144,148,163–166 Also, the neurogenic properties of the non-structural genes of YF 17D virus are somewhat attenuated in the DENV-YF chimeras.42,125,167 The impact of dengue vaccine administration on immunocompromised, immunosuppressed, or malnourished hosts is unclear. Severe DENV infections have not been reported in persons infected with human immuno-deficiency virus.168 These questions will be answered from the databases generated as vaccine candidates progress into large-scale clinical trials. The risk of recombination events leading to virulent reversion with the use of live virus vaccines has been proposed and carefully refuted.169–173 A more difficult issue concerns the long-term safety of chimeric viruses. Only pro-longed observations can satisfactorily resolve this issue.174 The recent licensure of Sanofi Pasteur's Japanese encephalitis vac-cine (ChimeriVax-JE; JE-CV; IMOJEV) shows that regulatory hurdles for genetically engineered vaccine candidates may be overcome.175

Vectored vaccines

Recombinant poxviruses and adenoviruses expressing foreign proteins have been demonstrated to induce strong humoral and cellular responses in humans against various pathogens.176–178 These viruses can infect cells and express their proteins de novo in the cell. The antigens are then naturally processed, glyco-sylated, and associated with the cell membrane. MHC class I–dependent immune responses are induced as a result of intra-cellular translation and processing of the gene products.

Vaccinia virus vector

Use of recombinant vaccinia viruses to express the structural proteins of DENV-2 or DENV-4 have been disappointing.179,180 These constructs, expressing prM and a full-length E protein, failed to induce neutralizing antibodies or to protect monkeys from wild-type challenge. Recombinants expressing truncated DENV E proteins did raise neutralizing antibody in mice and in rhesus macaques, providing some protection against wild-type challenge in both models.181

Adenovirus vector

Replication-defective adenovirus vectors have been used to construct flavivirus vaccine candidates.182,183 A DENV vaccine candidate was generated by inserting the E ectodomain gene of DENV-2 NGC into region 1 of the adenovirus 5 genome. Immunized mice developed type-specific B- and T-cell responses and anti-DENV-2 neutralizing antibodies.184 In another study, a recombinant adenovirus capable of expressing the E domain III (EDIII) of DENV-2 was created and tested in combination with a plasmid encoding the same domain. Heterologous (Ad/plasmid) prime-boost regimens were explored and revealed that both regimens (Ad → plasmid; plasmid → Ad) induced neutral-izing antibodies.185,186 A bivalent vaccine candidate was made by taking a recombinant, replication-defective adenovirus (rAd) vector encoding a chimeric antigen made of in-frame linked EDIIIs of DENV-2 and DENV-4. As in the monovalent study, a plasmid vector encoding the same chimeric bivalent antigen was used in a prime-boost strategy eliciting balanced neutraliz-ing antibody and specific T-cell responses.187

All four dengue genes have been used to make a successful tetravalent vaccine in animal models.188 Other investigators used a similar approach to construct a bivalent DENV-1–DENV-2 candidate.189 Subsequently, to ensure that the prM and E anti-gens are expressed at similarly high levels, the genes from each DENV were placed into identical, immediate-early cytomegalo-virus (CMV)–bovine growth hormone polyadenylation cassettes and inserted into either end of the adenovirus (cAdVax) vector. DENV-1 and -3 were injected into one arm and DENV-2 and -4 into the other.189,190 Both constructs induced neutralizing anti-bodies against all four DENV serotypes and a cellular immune response was detected 4-10 weeks following primary vaccina-tion.190 Investigators subsequently evaluated the two recombi-nant adenovirus vectors given together as a two-dose tetravalent DENV vaccine in rhesus monkeys and demonstrated significant protection against DENV challenge.191 This approach of a four-in-one physical mix has been bypassed by inserting a single tet-ravalent gene of the domain III region of all four DENV. This virus produced cell-mediated immune and neutralizing antibody responses in mice given four doses of vaccine.188 Of interest, this replication-defective adenovirus 5 vaccine construct raised immune responses in adenovirus 5–immune mice.

Venezuelan equine encephalitis virus vector

Nonpropagating Venezuelan equine encephalitis virus (VEE) replicon particles (VRP) permit translation and expression of inserted genes during a single round of replication.192–194 Monotypic DENV-1 and -2 vaccines have been constructed using VEE VRPs.195,196 In a prime-boost regimen, after two doses of a DNA vaccine, a single dose of VEE VRP resulted in complete protection of cynomolgous macaques.195 It has been possible to use VEE VRPs to overcome maternal-antibody interference in a mouse model.196 Unlike live attenuated vaccines that immu-nize as a result of multiple rounds of replication, nonpropa-gating VRP vectors translate high levels of heterologous genes during a single round of infection. Rhesus macaques have been immunized with 108 international units of DENV3Es-VRP at


weeks 0, 7, and 18. All animals produced neutralizing antibod-ies and showed complete protection from viremia when chal-lenged with live virus (L. White, personal communication, 2010). A tetravalent Es-VRP has been constructed, and testing in nonhuman primates is planned. Of interest, VEE glycopro-teins appear to target lymph node dendritic cells, resulting in an adjuvant activity. 197 , 198

Measles virus vector

Genes for the EDIII of DENV-1 fused to those of the ectodo-main of the membrane protein (ectoM) were inserted into the Schwarz measles vaccine (MV) strain. 199 Measles-susceptible mice inoculated intraperitoneally with two doses of this con-struct produced DENV-1 neutralizing antibodies. Mice were challenged with live DENV-1 six months after the initial immunization. The presence of ectoM was critical to the immunogenicity of inserted EDIII. The adjuvant capacity of ectoM correlated with its ability to promote the maturation of dendritic cells and the secretion of proinflammatory and anti-viral cytokines and chemokines involved in adaptive immunity. Subsequently, a recombinant MV vector containing genes for all four DENV EDIII fused with ectoM was prepared. 200 After two injections in mice susceptible to MV infection, the recombinant vector induced neutralizing antibodies against the four DENV types. Curiously, the authors interpreted anamnestic responses observed when immunized mice were challenged with each of the four DENV as evidence of protection. However, in carefully studied dengue virus challenges of immunized rhesus mon-keys, anamnestic neutralizing antibody responses were inter-preted as evidence of in vivo replication of the challenge virus and therefore as partial vaccine failure. 64 “Protected” animals responded to challenge with only a brief, low-level hemagglu-tination-inhibition antibody rise but no change in neutralizing antibody titers.

Replication-defective vaccines

Capsid-gene deletion

Several flavivirus vaccines have been made by deleting the cap-sid gene (C) and propagating the virus in cells that express high levels of C. 201 Such a virus undergoes only a single cycle of infec-tion in vaccinated hosts. A DENV-2 RepliVAX, produced by replacing the prM/E genes of RepliVAX West Nile virus with the same genes of DENV-2, resulted in partial protection of AG 129 mice challenged with a mouse-adapted DENV 2. 202

Inactivated or subunit vaccines

The simplest approach to produce viral vaccines is the use of inactivated whole virus or viral subunits. Such vaccines have potential advantages—for example, they cannot revert to a more pathogenic phenotype, and when combined they are unlikely to produce immune interference. Thus, they may be suitable for a wider age range than live vaccines, immunosup-pression would not be a contraindication to immunization, an accelerated immunization schedule may be possible, and, theo-retically, there could be fewer safety issues. Cell-mediated as well as humoral immune responses have been demonstrated with an inactivated flavivirus vaccine. 203

On the other hand, killed or subunit vaccines raise anti-bodies to only a portion of the structural proteins and normal virion-based structural conformation. Other disadvantages include the requirement for high concentrations of antigens and the need for multiple doses. Inactivated flaviviral vaccines have been licensed and are in wide use to prevent Japanese

encephalitis 204 , 205 and tick-borne encephalitis. 206 Inactivated vaccines against flaviviral encephalitis may succeed because immune enhancement is not part of the disease process.

Whether inactivated or subunit dengue vaccines can be used as standalone products is debatable. They may be useful for relatively short-term protection or in a prime-boost strat-egy. Augmentation of inactivated vaccines with new proprie-tary adjuvants may improve manufacturing and performance parameters. 207 Recent studies on respiratory syncytial and mea-sles viruses have shown that formalin not only de-conforms viral proteins, resulting in poor B-product and T-cell func-tioning, but also interferes with normal processing by den-dritic cells. 208 , 209 This result suggests that other, less stressful mechanisms of inactivating DENV should be sought. Psoralen-inactivated DENV 1 was used to immunize in Aotus monkeys but with only equivocal protection on challenge. 210

Cell culture–based inactivated vaccines DENV-2 (S-16803) grown in Vero cells, concentrated by ultracentrifugation, and purified on a sucrose gradient prior to formalin inactivation and administration with alum suc-cessfully protected rhesus monkeys in experiments conducted by the WRAIR. 211 , 212 Further monkey challenge studies were performed on this purified, inactivated vaccine (D2-PIV [puri-fied, inactivated vaccine]), which was formulated with four different adjuvants (alum and GlaxoSmithKline adjuvants SBAS4, SBAS5, and SBAS8) (see Table 44-1 ). 40 Head-to-head comparisons were made with LAV DENV-2, S16803 PDK-50. Two doses of D2-PIV, D2 LAV, or saline (control) were given 3 months apart, and animals were challenged with the D2 S16803 parent 3 months after the second dose. All but one vaccinated animal seroconverted after the first dose of vac-cine, with anamnestic antibody rises after the second dose. After virus challenge, viremia was consistently observed in controls, and sporadically in D2-PIV/alum and SBAS4 groups, but not in animals that received D2-PIV/SBAS5, D2-PIV/SBAS8, or D2 LAV. Some animals that were rechallenged with wild-type virus 1 year later were protected against viremia. 213 GlaxoSmithKline Biologicals are leveraging their proprietary adjuvant systems and collaborations with the WRAIR and the Oswaldo Cruz Foundation—Fiocruz to explore further development of the PIV concept pioneered by the WRAIR. 214 Although the benefit of using adjuvant with flavivirus vac-cines is not yet studied, dosing requirements, safety profile, immunogenicity, and time to protection could all benefit from combining an inactivated dengue vaccine with an adjuvant. Use of a prime-boost strategy is another alternative to achieve solid protection. Rhesus monkeys vaccinated with inactivated tetravalent dengue vaccine and then boosted 2 months later with live attenuated vaccine developed high levels of tetra-valent dengue neutralizing antibodies and resisted live virus challenge. 215 , 216

Recombinant subunit vaccines T- and B-cell epitopes have been mapped on DENV structural and nonstructural proteins. 217 , 218 The right combination of epi-topes expressed in protein subunit vaccines might be the basis for an effective vaccine produced at moderate cost. 219 Structural and nonstructural DENV proteins can be produced in large quantities in many expression systems including Escherichia coli, 220–222 baculovirus in Spodoptera frugiperda insect cells, 223–226 yeast, 227 vaccinia virus, 180 , 228 , 229 and Drosophila cells. 40 The last approach, which resulted in 80% E gene expression in Drosophila cells, is being developed by the Hawaii Biotech Inc/Merck & Co. (see Table 44-1 ).

A tetravalent vaccine has been developed consisting of 80% E proteins of each of the four dengue viruses. Efficacy expectations are based on experiments showing that 1 μ g of


DENV-2 E protein with the SBAS5 adjuvant protected rhe-sus monkeys from viremia after challenge with wild-type virus.213 Early development included a DENV-2 vaccine can-didate with 80% E + nonstructural protein 1 (NS1), but the NS1 added little to performance and this candidate has been sidelined. Various formulations have been tested in rhesus macaques producing high-titer neutralizing antibodies to all four dengue viruses and protecting animals from viremia after wild-type viral challenge.40 A phase 1, single-center, double-blind, randomized study to assess the safety and tol-erability of a DENV-1 candidate (HBV-001 D1) in healthy US adults was recently completed. High and low dosages (50 μg and 10 μg of DEN1-80E) were tested against placebo in 16 subjects across two cohorts. Three doses of DEN1-80E (HBV-001 D1 + 3.5 mg Alhydrogel) vaccine (0.5 mL) were admin-istered intramuscularly at 0, 4, and 8 weeks; publication is pending results.230

Cuban investigators have produced candidate dengue vac-cines using recombinant proteins. DENV-3 DIII protein fused with a P64k Neisseria meningitidis carrier protein, adminis-tered with Freund's complete adjuvant, protected nonhuman primates after challenge.231 However, only partial protection was obtained using aluminum hydroxide as the adjuvant, and this required use of the Neisseria serogroup A polysaccharide.232 The authors suggest priming with dengue protein and boosting with live attenuated virus.233

Synthetic peptidesSynthetic peptides containing B- and T-cell epitopes are immu-nogenic in mice, and combinations of peptides have been tested as subunit vaccines.234,235 Antibodies directed to syn-thetic peptides have been detected in sera from patients con-valescing from dengue infections.236,237 Peptides are unlikely to raise antibodies that will completely protect against DENV infection as they may provide fewer of the conformational epi-topes required.

Exploring the safety of inactivated vaccinesThe use of inactivated, subunit, or peptide vaccines is expected to reduce the problems of reactogenicity, interference, and pos-sible antibody-enhanced worsening of vaccine virus reactoge-nicity posed by live virus vaccine candidates. Chief concerns for inactivated vaccines are their reliance on antibodies to medi-ate protection and, absent use of newer adjuvants, the possible short duration of high titers of protective levels of neutralizing antibodies after the administration of these vaccines. As is evi-dent from the wide taxa of viral diseases of nonhuman species where Fc receptor–bearing cells are principal target cells for viral infection, inactivated or subunit vaccines that raise nonpro-tective antibodies often result in breakthrough infections and enhanced disease.24,238

The precise mechanism regulating antibody responses after virus infections is not fully understood. Formalin inactivated and subunit vaccines may lack the structural configurations of viral proteins crucial to eliciting the innate and adaptive immune responses that result in complete protection. For some human vaccines, formalin-inactivated whole-virus prep-arations have led to well-studied vaccine-induced immunopa-thologies.239 For example, nonprotective complement-fixing antibodies with low affinity for measles virus resulted in atyp-ical measles.240,241 An analogous outcome has been observed with respiratory syncytial virus (RSV). There is evidence that the use of formalin de-conforms viral epitopes of RSV, resulting in poor Toll-like receptor stimulation; suboptimal maturation of dendritic cells with reduced production of activation fac-tors CD40, CD80, and CD86; decreased germinal-center for-mation in lymph nodes; and the production of nonprotective

antibodies. These antibodies fail to neutralize RSV, allowing replication with secondary stimulation of RSV-primed Th2 cells producing more low-avidity antibody, resulting in com-plement-activating immune complexes being deposited into affected tissue.208 When Toll-like receptor agonists were incor-porated into a formalin-inactivated RSV vaccine (lipopoly-saccharide, ss poly I:C and ss polyU), it was possible to raise protective immune responses in mice.208

For nonflaviviruses, adjuvants, particularly newer complex adjuvant systems (AS), have been shown to mediate long-lasting antibody responses. Two, AS03 and AS04, have been incor-porated into vaccines licensed for human use.242–244 Although experimental confirmation is lacking in the Flaviviridae fam-ily, the use of the adjuvant systems is being explored by the WRAIR/GSK/Oswald Cruz Foundation Killed Dengue Vaccine Initiative.

Nucleic acid–based vaccines

DNA vaccines consist of a plasmid (or plasmids) containing DENV genes reproduced to high copy number in bacteria such as E. coli.245 The plasmid contains a eukaryotic promoter and termination sequence to drive transcription in the vaccine recipient. The transcribed RNA is translated to produce pro-teins to be processed and presented to the immune system in the context of MHC molecules. Additional genes such as intra-cellular trafficking and immunostimulatory sequences can be added to the plasmid. Expressed antigens lead to B- and T-cell responses. DNA vaccines afford numerous theoretical advantages over conventional vaccines, including ease of pro-duction, stability and ability to be transported at room tem-perature, ability to add new genes to the vaccine, ability to immunize against multiple pathogens with a single construct, and reduced reactogenicity.

PrM-E DNA vaccines

Workers at the Naval Medical Research Center evaluated two eukaryotic plasmid expression vectors (pkCMVint-Polyli and pVR1012; Vical, Inc., San Diego, CA) expressing the PrM pro-tein and 92% of the E protein for DENV-1 and DENV-2 virus (see Table 44-1). These constructs induced neutralizing anti-body in mice,246 and they were subsequently improved by add-ing immunostimulatory CpG motifs, and the full-length E gene with PrM.247–250 In recent studies, a DENV-1 DNA vaccine pro-tected a portion of challenged monkeys from viremia for varying lengths of time.251 With the recognition that dendritic cells are the percutaneous portal for dengue virus replication,252 efforts are being made to target DNA vaccines to these cells.213 A DENV-1 DNA vaccine was administered intramuscularly to 22 flavivirus-negative volunteers, half of whom received high and low dosages, respectively, in a three-dose series (at day 0, and at 1 and 5 months). At the completion of this series, none of the low-dosage recipients and only five of 11 high-dosage recipients developed neutralizing antibodies.253 More recently, protection has been achieved in a rhesus monkey model by boosting tetra-valent DNA vaccination with a tetravalent live attenuated virus vaccine.215

Tetravalent dengue vaccines have been created by shuffling the envelope genes from the four dengue viruses. Selected shuf-fled DNA was transfected into human cells, subjected to flow cytometry, and reacted with type-specific dengue antibodies. Antibody markers permitted rapid screening of libraries and identification of novel expressed chimeric antigens. A panel of chimeric clones expressing C-terminal truncated antigens that combined envelope and premembrane epitopes from all four


DENV types when inoculated in mice and monkeys success-fully raised neutralizing antibodies. Monkeys resisted challenge with DENV-1 but not DENV-2. 254 , 255

Exploring the safety of DNA vaccines

Although the DNA approach offers advantages, it also carries unique risks. 256 These include the theoretical risk of nucleic acid integration into the host's chromosomal DNA to potentially inac-tivate tumor suppressor genes or to activate oncogenes. This risk appears to be well below the spontaneous mutation frequency for mammalian cells. 257 , 258 However, if a mutation resulting from DNA integration is a part of a multiple hit phenomenon lead-ing to carcinogenesis, it could be many years before this problem becomes evident. Another concern is that foreign DNA might induce anti-DNA antibodies, leading to autoimmune diseases such as systemic lupus erythematosus. Studies in lupus-prone mice, normal mice, rabbits, and people have not validated this concern, 259 , 260 and, in fact, DNA vaccines are being proposed as an approach to the management of autoimmune diseases. 261 , 262

Conclusions

Dengue virus infections occur regularly among military per-sonnel assigned to combat or peacekeeping roles in tropical countries. These infections are consistently among the lead-ing causes of febrile diseases in tourists and expatriate residents of tropical countries. Considerable morbidity and mortality is inflicted on native populations of all social strata in endemic areas. These medical and social impacts create a significant market, as evidenced by the interest shown by major vaccine manufacturers. As described here, several promising dengue vaccine candidates are in preclinical and clinical development, and one manufacturer has entered phase 3 testing. If the safety concerns described can be surmounted, economic forces and technologic advances should soon bring one or more dengue vaccines to the market. It remains for the vaccine community to develop and implement plans for the strategic use of dengue vaccines by developing evidence-based policies to target high-risk groups and decrease virus transmission.

Access the complete reference list online at http://www.expertconsult.com

3 . Guzman MG , Halstead SB , Artsob H , et al . Dengue: a continuing global threat . Nat Rev Microbiol 2010 ; 8 (12 Suppl.) : S7 – 16 .

5 . Lum LC , Suaya JA , Tan LH , et al . Quality of life of dengue patients . Am J Trop Med Hyg 2008 ; 78 : 862 – 7 .

10 . Suaya JA , Shepard DS , Siqueira JB , et al . Cost of Dengue cases in eight countries in the Americas and Asia: a prospective study . Am J Trop Med Hyg 2009 ; 80 : 846 – 55 .

11 . Shepard DS , Coudeville L , Halasa YA , et al . Economic impact of dengue illness in the Americas . Am J Trop Med Hyg 2011 ; 84 : 200 – 7 .

23 . Halstead SB . Pathogenesis of dengue: challenges to molecular biology . Science 1988 ; 239 : 476 – 81 .

29 . Rothman AL . T lymphocyte responses to heterologous secondary dengue virus infections . Ann N Y Acad Sci 2009 ; 1171 (Suppl. 1) : E36 – 41 .

51 . Swaminathan S , Batra G , Khanna N . Dengue vaccines: state of the art . Expert Opin Ther Pat 2010 ; 20 : 819 – 35 .

52 . Murphy BR , Whitehead SS . Immune response to dengue virus and prospects for a vaccine . Annu Rev Immunol 2011 ; 29 : 587 – 619 .

238 . Halstead SB , Mahalingam S , Marovich MA , et al . Intrinsic antibody-dependent enhancement of microbial infection in macrophages: disease regulation by immune complexes . Lancet Infect Dis 2010 ; 10 : 712 – 22 .

vaccines || dengue vaccines

Documents