review

Human Vaccines 6:9, 696-705; September 2010; © 2010 Landes Bioscience

Development of Sanofi Pasteur tetravalent dengue vaccine Bruno Guy,* Melanie Saville and Jean Lang Research and Discovery Departments; sanofi Pasteur; Marcy l’Etoile, France

Key words: dengue, vaccine, human, development, immunity

The Sanofi Pasteur tetravalent dengue vaccine candidate is composed of four recombinant live attenuated vaccines based on a yellow fever vaccine 17D (YFV 17D) backbone, each expressing the prM and envelope genes of one of the four dengue virus serotypes. Pre-clinical studies have demonstrated that the TV dengue vaccine is genetically and phenotypically stable, non-hepatotropic, less neurovirulent than YFV 17D and does not infect mosquitoes by the oral route. In vitro and in vivo preclinical studies also showed that the TV dengue vaccine induced controlled stimulation in human dendritic cells and significant immune responses in monkeys. TV dengue vaccine reactogenicity, viramia induction and antibody responses were investigated in three Phase I trials in the USA, the Philippines and Mexico, in a two or three-dose regimen over a 12 month period. Results showed that the majority of adverse events were mild to moderate and transient in nature. Viremia was transient and low, and was not increased after initial dengue TV administration, even in the case of incomplete responses. fSeropositivity (≥10 in a PRNT 50 assay) was 100% for all four serotypes in flavivirus-naive adults injected with 3 doses of TV dengue vaccine in the USA. Similarly, seropositivity was 88–100% following three administrations in flavivirus-naive Mexican children aged 2–5 years. Furthermore, the proportion of seropositive subjects increased with each dengue TV injection in the Philippines where baseline flavivirus immunity was high. Responses were also monitored at the cellular level in humans, and their level and nature were in good agreement with the observed safety and the immunogenicity of the vaccine. Finally, the challenges inherent to the development of such TV dengue vaccines will also be discussed in the last part of this review. In conclusion, preclinical and clinical results support the favorable immunogenicity and short-term safety of the dengue TV vaccine. An extensive clinical development program for dengue TV is underway including completion of the enrollment of 4,000 4–11 years old children in an efficacy trial in Thailand, in an area of high dengue incidence. Assuming continued successful outcomes, initial submissions to regulatory authorities are envisaged within a 5-year period.

*Correspondence to: Bruno Guy; Email: [email protected] Submitted: 05/28/10; Accepted: 06/16/10 Previously published online: www.landesbioscience.com/journals/vaccines/article/12739 DOI: 10.4161.hv.6.9.12739

696

Introduction Several members of the Flavivirus genus are serious threats to human and animal health. Among them, dengue viruses represent a major and growing medical problem. All of the four serotypes of dengue virus can cause clinical manifestations ranging from selflimiting dengue fever to severe dengue hemorrhagic fever (DHF) and fatal dengue shock syndrome (DSS). The number of dengue infections in endemic areas has continued to increase over the past two decades. Over one hundred countries are affected with over three billion people at risk (reviewed in ref. 1). It has been estimated that more than 100 million dengue infections resulting in 24,000 deaths occur annually. Children are the most affected. This increased disease incidence and extended geographical reach of dengue2 have made the development of an effective vaccine an international health priority. Academic laboratories and pharmaceutical companies have developed several dengue vaccine candidates using a variety of technologies, including live-attenuated vaccines (LAVs), recombinant virus vectors expressing dengue envelope (E) antigens, recombinant proteins and DNA vaccines, none of which have been licensed (for reviewes see refs. 3–5). Sanofi Pasteur is now developing a tetravalent live attenuated chimeric dengue virus vaccine based on the yellow fever 17D vaccine strain. The technology (sometimes referred to as ChimeriVax) behind the production of the chimeric dengue vaccine viruses originated at St. Louis University,6 and was applied by Guirakhoo et al. at Acambis Inc., (now a part of Sanofi Pasteur).7 As we will develop in this review, this dengue vaccine candidate is immunogenic and safe in humans, and is currently being evaluated in large scale efficacy studies. The live attenuated and chimeric nature of these vaccine viruses necessitates extensive preclinical and clinical characterization, from the early stages of research through to clinical development. Furthermore, their status as genetically modified organisms (GMO) required compliance with additional specific regulations. These issues were reviewed in reference 8, focusing on the various tools, assays and research strategies implemented as part of our chimeric flavivirus research and development program, and will be discussed only briefly here. Construction of the Tetravalent Dengue Vaccine The four monovalent chimeric vaccine viruses, CYD1-4, were constructed by replacing the genes for YF vaccine (YFV 17D 204)

Human Vaccines Volume 6 Issue 9

Product REVIEW

review

Figure 1. Construction of the four chimeric vaccines.

premembrane (prM) and envelope (E) proteins, with those of the four dengue serotypes6,7,9 (Fig. 1). A tetravalent CYD1-4 dengue vaccine (TDV) is produced by combining the four monovalent viruses in a single vaccine preparation. Preclinical Evaluation Before entering clinical trials, all vaccine candidates must be tested for safety and immunogenicity in preclinical studies. For these dengue vaccines, testing can be conducted both in vitro on primary or transformed cells, including human cells, and in vivo in animals, in particular non-human primates (NHP). Preclinical studies were designed to provide information on the phenotypic and genotypic stability of the vaccine candidates, their tropism, structure, ability to replicate and to be transmitted by mosquito vectors, as well as to document specific aspects linked to the use of genetically modified organisms (GMOs). All of the above directly or indirectly affect safety and immunogenicity. It is also important that a dengue vaccine provides protective immunity against all four circulating viral serotypes, a point that was also addressed in preclinical studies. Figure 2 presents the different types of cells and tissues potentially affected by dengue vaccines and Figure 3 shows some types of preclinical studies that can be carried out to study the consequences of such interactions.

www.landesbioscience.com

In vitro genetic stability. The YFV 17D vaccine genome is remarkably stable, both in vivo and in vitro,10 which may be attributed to the low error-rate of the virus RNA polymerase.11 The same enzyme assures the viral replication of the CYD viruses and it was thus expected that these viruses would display a similarly high stability. The full genome sequence of the each CYD was established at various stages during the manufacture of GMP vaccine lots, from the first passages, to premaster seed lots (PMSL), master seed lots (MSL) and bulk and ultimately at a later step in the process (bulk +10 passages). Each of the four CYD viruses exhibited high genomic stability in cell culture, with a total of only nine mutations observed. Five of these mutations were detected only at late passages (between p10 and p21), and three of them found in a mixed population with the original sequence. One mutation was silent. All mutations except one in the E gene were located in the non-structural regions of the genome and likely reflect adaptation to Vero cells. The non-silent mutations present at early manufacturing steps appeared during the scale change between PMSL to MSL and were then conserved stably throughout the process (unpublished data). In vitro phenotypic stability. Consensus genome sequencing is unsuitable for the detection of minor, quasi-species in a vaccine seed or batch and gives little information about the potential biological consequences of a particular mutation. However,

Human Vaccines

697

Figure 2. Innate and adaptive immune responses potentially triggered by dengue vaccine candidates. Upon vaccination, dengue vaccine viruses may replicate preferentially in cells from the monocytic lineage, such as monocytes and dendritic cells (DCs). Antigen presentation to CD4/CD8 T cells will trigger their activation and subsequent B-cell activation. Activated memory B cells can also constitute potent antigen-presenting cells upon boosting. It is also of interest to monitor the potential replication and consequences, of dengue vaccines in other cell types, including endothelial cells and hepatic cells.

mutations that affect the infection efficiency, growth, penetration or spread of the virus in cell culture, generally modify the plaque phenotype.12,13 Phenotype consistency was therefore monitored throughout vaccine lot production using a plaque size phenotype assay. By measuring at least 100 plaques, it was estimated that we achieved a greater than 90% probability of detecting revertants representing 2% of the total population. We found the plaque size phenotype of all four CYD viruses to be stable at all production steps. Phenotypic stability was also assessed in animal models when available (see below). In vitro preclinical studies. Skin dendritic cells (DCs) are among the first cells to encounter virus after inoculation and are also the most efficient antigen-presenting cells (APC) implicated in the primary immune response. Interactions between human DCs and wild-type (wt) dengue viruses have been well documented,14,15 it was therefore interesting to compare immune consequences of infection with attenuated vaccine viruses versus their wt parents. We investigated the infectivity of the four CYD viruses in monocyte-derived human DCs,16 as well as the consequences of infection in terms cellular activation and maturation and the secretion of pro- and anti-inflammatory cytokines, chemokines and type I interferons.17 The CYD1-4 viruses were seen to induce DC maturation and a controlled response, accompanied by limited inflammatory cytokine production and consistent expression of anti-viral type I IFN, in agreement with

698

their good clinical safety profile and immunogenicity (see below). These results were further confirmed and extended using DNA array profiling (Guy B et al. J Infect Dis; In Press). The tropism of flaviviruses is mostly linked to the envelope, so CYD should not exhibit the same tropism as YFV 17D. To test this hypothesis, the growth kinetics of CYD1-4 and parental viruses (wild-type DEN-1-4 and YF 17D) were assessed in three hepatic cell lines—HepG2, Huh7 and THLE-3—as a potential marker of viscerotropism. Replication of all chimeric viruses in HepG2 and THLE 3 cells, but not in Huh7, was markedly lower than that of YF-VAX.16 Differences in findings between cell lines may be explained by the fact that Huh7 cells are permissive to replication of many viruses, irrespective of their attenuated phenotype. These results nevertheless suggest that the CYD1-4 viruses are less hepatotropic that YF 17D virus vaccine in humans. Other assays which will not be described here have also been used to further characterize the CYD vaccine candidates. These include: electron microscopy to assess viral maturity; SDS/PAGE analysis to assess the consistency of the protein content and profile of the vaccines, glycosylation status; replication in DC SIGNtransfected cell lines to assess the ability of vaccine candidates to interact with this molecule, and subsequently effectively enter cells; replication in insect C6/36 cells and temperature sensitivity assays.

Human Vaccines Volume 6 Issue 9

Figure 3. Preclinical assays used to characterize dengue vaccine candidates.

In vivo preclinical studies. Mouse models of neurovirulence have been used to discriminate between the neurotropism of dengue vaccine candidates with that of their parental viruses. One such model in suckling mice has been shown to be an acceptable alternative to NHP models.18 After i.c. inoculation in both mice and NHPs, all four CYD viruses were significantly attenuated, even compared with the parental YFV 17D vaccine. The suckling mouse model is now routinely used for in-process control testing during the manufacturing process of Sanofi Pasteur’s YFV 17D-based flavivirus vaccines. Some NHPs, including rhesus (Macaca mulatta) and cynomolgus monkeys (Macaca fascicularis), are sensitive to dengue and YF infections. The World Health Organization (WHO) recognizes these species as good models for the assessment of the neurotropism and the viscerotropism of attenuated YF vaccines (WHO Technical report series, N° 872, 1998). Viremia can be used to assess the attenuation of vaccine candidates by comparing vaccine virus viremia with that of the wt parental strains. As dengue-infected monkeys do not develop disease symptoms, viremia can also be used to assess protection against wt viral challenge by evaluating the reduction in wt virus viremia in vaccinated animals compared with unvaccinated control animals. Viremia can thus be considered as both a direct indicator of tropism and an indirect indicator of safety since it has been identified as one of the factors associated with virulence and disease severity in humans.19 NHP studies can also be used to provide additional information on the ability of dengue vaccine candidates to elicit

www.landesbioscience.com

neutralizing antibodies.9,20,21 In particular they have been used to explore immunization parameters such dosing regimen and the vaccine formulation, and to monitor for viral interference between the four vaccine viruses. As absolute thresholds for protective immunogenicity and acceptable viremia are difficult to establish, and cannot necessarily be extrapolated between host species, immune responses and viremia induced by vaccine candidate viruses are compared with those induced by wt viruses. NHP studies have shown that primary immunization with TDV induced a short-lived, low-level viremia, which was not present after booster immunizations. One or several injections of TDV conferred immunity against the four wt dengue serotypes, and almost complete protection against each serotype upon subsequent wt challenge.9 As with any multivalent vaccine, dengue vaccine development is complicated by the potential for interference between serotypes which can result in a dominant immune response against only one or two serotypes. In NHPs, interference was observed after vaccination with a tetravalent preparation containing 5 log10 CCID50 of each virus (designated 5,555, and considered as the reference tetravalent vaccine preparation).22 We identified several approaches to mitigating interference: (1) simultaneous administration of two complementary bivalent vaccines at separate anatomical sites drained by different lymph nodes; (2) sequential administration of two complementary bivalent vaccines; (3) preimmunization with a heterologous flavivirus; (4) adaptation of formulations by decreasing the dosage of the immunodominant

Human Vaccines

699

Figure 4. GMT (log10 with 95% CI, WHO reference strains) in naïve volunteers obtained after 1 to 3 immunizations of TV dengue vaccine (from ref. 32). Group 1: 3 doses of TDV given at D0, M4 and M12. Group 2: one dose of placebo given at D0 followed by 2 doses of TDV given at M4 and M12. Blood samples were taken prior to each dose of vaccine and 28 dose post vaccination.

serotype 4 CYD virus and (5) administration of a booster at 1 year. These studies also showed that immunizations should be spaced several months apart to prevent negative interference, possibly due to short-lived cross-reactive (IgM) antibodies, crossreactive T cells or to innate immunity, and to allow a better induction of memory. Such regimens have been tested in humans in different trials and have confirmed the importance of a one year booster. These studies have also highlighted the differences between species, as the interval between sequential immunization with complementary bivalent vaccines needs to be longer in humans than in monkeys (unpublished data). Environmental Risk Assessment A number of theoretical issues associated with the live and GMO nature of the CYD viruses have been addressed throughout their development as reviewed in detail in reference 8. The four most frequently raised concerns are: transmission by arthropod vectors, recombination with a circulating virus, reversion to virulence and risks of viscerotropism.

700

To assess whether mosquitoes could become infected with the CYD viruses from feeding on a vaccinated host, the ability of CYDs to replicate in Aedes albopictus mosquito cell culture (C6/36) and in Ae. aegypti mosquitoes, the principal vectors of YF and dengue viruses, was evaluated in comparison with the parental YF 17D and wt dengue viruses.23 The CYD viruses were unable to infect orally Ae. aegypti and Ae. Albopictus or to replicate in midgut tissue after intra-thoracic inoculation, and were more attenuated than YF 17D virus in these species. Together with the absent or low-level, short-lived CYD viremia in human vaccinees, the inability to infect and replicate in mosquitoes vectors are critical safeguards against the dissemination of CYD viruses in the environment. Based on theoretical assumptions and an analogy with nonflavivirus vaccines, it was suggested that new viruses might emerge from the recombination between flaviviruses or with dissimilar RNA viruses,24 although the authors’ assumptions were challenged by most experts in the field.25,26 A recent study investigated the likelihood of intermolecular recombination between different flaviviruses using pairs of replicons derived

Human Vaccines Volume 6 Issue 9

from Tick-borne encephalitis virus (TBVE), Japanese encephalitis virus (JEV) and West Nile virus (WNV).27 The very few recombination events detected (none for TBEV or WNV), were aberrant recombinations resulting in virus with impaired growth properties. Results showed that flaviviruses have a low propensity for homologous recombination. The findings complement those of previous studies investigating what would happen in the unlikely event that such a recombinant virus emerged. We considered the worst-case scenario where one of the attenuated CYD vaccine viruses recombined with the wt YF Asibi virus. All observations of replication and transmission in mosquitoes, as well as outcomes in NHPs showed that these recombinants were attenuated compared with the parental wt viruses.28,29 These studies also suggested that the chimerization process itself contributed to the attenuation of these viruses. Thus, not only is the recombination of the CYD vaccine viruses with a wt flavivirus extremely unlikely, any recombinant would be unlikely to cause disease or be disseminated. While reversion to virulence has been raised by some as a potential concern, reversion of a CYD virus into a virulent YF virus is virtually impossible given the genetic stability discussed above. In addition to not having YFV 17D preM or E genes, the CYD chimeric virus features numerous attenuating residues within the seven YFV 17D nonstructural genes and the capsid protein gene. Reversions in all of these would be required for virulent virus to emerge. Although extremely rare, acute viscerotropism disease has occurred following YF 17D vaccination (estimated incidence: 0.3–0.4 per 100,000 vaccinated individuals).30 There is a perceived risk that such a serious adverse event might therefore occur after CYD vaccination. However, as stated previously, the viral tropism and virulence are largely linked to the E protein, and the E gene is precisely one of the YFV 17D genes that is not present in the chimeric dengue viruses. Additionally, as described above, in vitro and preclinical in vivo experiments have shown that the viscerotropism and neurotropism of CYD viruses are significantly attenuated compared with YFV 17D. It is thus plausible to propose that the safety profile of chimeric dengue vaccines will be improved over that of YFV 17D, particularly with respect to most neurotropic adverse events. Clinical Development The considerations and challenges of clinical development include: (1) the need to induce an adequate and balanced immune response to all four serotypes; (2) the need for two or three vaccinations over a period of up to 12 months in flavivirus-naïve individuals; (3) the current absence of a correlate and threshold of protection and thus the need to demonstrate clinical efficacy; (4) the need to demonstrate long term safety and immunogenicity; (5) the theoretical risks of sensitization to severe dengue infection (DHF) after vaccination and of acute viscerotropic disease (AVD) and neurotropic disease (AND) that are very rare serious adverse events after YFV 17D vaccination and (6) the need to comply with GMO regulations. Additional complexity is brought by the fact that the flavivirus immunological background

www.landesbioscience.com

varies between regions, as do the current recommended vaccination schedules, meaning trials will have to be conducted in parallel on different continents potentially with different coadministered vaccines. In addition to compliance with Good Clinical Practice guidelines and all applicable national and international regulations, clinical trials are being designed and conducted in accordance with the WHO Guidelines for the Clinical Evaluation of Dengue Vaccines in Endemic Areas (WHO/IVB/08.12). Indeed, our first priority is to develop the vaccine in endemic countries of Asia-Pacific, Latin America and the Caribbean to address the unmet medical need for children and adults. Trials will also be performed in non-endemic countries, for example in Europe and USA, for travelers and military personnel. Phase I clinical evaluation. The first clinical evaluation of a CYD vaccine candidate was with a monovalent serotype 2 CYD virus (CYD-2). The safety profile of CYD-2 was reported to be similar to the YF-17D control (YF-VAX TM) and transient lowlevel viremia was observed. Most participants seroconverted to dengue virus strain 16,681, and prior YFV 17D immunity was found to induced stronger, broader (cross-protective) and long lasting anti-dengue immune responses.31 As of April 2010 more than 4,800 volunteers have received at least one dose TDV (5 log10 CCID50 of each serotype), including children from 2 years of age to adults up to 45 years, in both dengue-endemic (Philippines, Thailand) and non-endemic (USA, Mexico City, Australia) areas. The first study showing complete seropositivity (≥10 in a PRNT 50 assay) against the 4 dengue serotypes was carried out in dengue-naïve US adults who received 3 doses in a 0–3.5–12 month regimen or 2 doses 8–9 months apart.32 Consistent with the TDV’s favorable safety profile (Table 1), viremia was low and was mainly CYD-4 after the first vaccination. After the second TDV vaccination, almost no viremia was detected by either RT-PCR or plaque assay (PA) for any serotype. After the second vaccination, more than 85% of vaccinees had no detectable CYD-1, 2 or 3 viremia. This finding has significant implications for the safety of this vaccine as is shows that while the first TDV vaccination did not elicit complete seroconversion to serotypes 1, 2 and 3, this did not induce sensitization to a second TDV vaccination in the presence of heterologous antiserotype 4 antibodies. All participants who received three TDV vaccinations seroconverted to all four WHO reference strains tested (Fig. 4). An increased immune response with an increased number of doses was apparent: both GMT and the percentage of seroconverted participants increased after each vaccination. The first vaccination induced a neutralizing humoral response mainly against serotypes 4 and 2, and to a lesser extent against 1 and 3. The second and third doses increased the percentage of seroconverted participants as well as the GMTs for all four serotypes, balancing the immune response across all four serotypes (GMT of 67, 538, 122 and 154 against serotypes 1, 2, 3 and 4 respectively). Moreover, there was a trend towards higher GMTs and higher seroconversion rates after 2 TVD doses in volunteers receiveing these doses 8–9 months rather than 3 months apart (especially for serotypes 1 and 3). This latter observation

Human Vaccines

701

Table 1. Adverse event (AE) observed after TDV administration in naïve volunteers (from ref. 32) V1 D0 Group 1 1st TDV (N = 33)

V2 M4

Group 2 Placebo (N = 33)

Group 1 2nd TDV (N = 30)

V3 M12 Group 2 1st TDV (N = 33)

Group 1 3rd TDV (N = 23)

Group 2 2nd TDV (N = 26)

n (%)

n (%)

n (%)

n (%)

n (%)

n (%)

Any Adverse Event

28 (84.8)

22 (66.7)

24 (80.0)

24 (72.7)

15 (65.2)

17 (65.4)

Any Adverse Reaction

27 (81.8)

20 (60.6)

21 (70.0)

21 (63.6)

10 (43.5)

17 (65.4)

Solicited Injection site reaction

6 (18.2)

5 (15.2)

11 (36.7)

8 (24.2)

4 (17.4)

7 (26.9)

Solicited Systemic reaction

26 (78.8)

19 (57.6)

20 (66.7)

18 (54.5)

10 (43.5)

16 (61.5)

Severe adverse event

6 (18.2)

2 (6.1)

1 (3.3)

4 (12.1)

0 (0.0)

0 (0.0)

Severe solicited reaction

5 (15.2)

2 (6.1)

1 (3.3)

4 (12.1)

0 (0.0)

0 (0.0)

Injection site

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

Systemic

5 (15.2)

2 (6.1)

1 (3.3)

4 (12.1)

0 (0.0)

0 (0.0)

confirmed observations in monkeys immunized with the same TDV, where marked increases in immune responses against all serotypes were seen when there was a 10 month interval between the second and third vaccinations, but not when this interval was only two months.22 Cellular immune responses were also monitored in this clinical trial. The observed level and nature (cytokine profile, CD8/Th bias, serotype dominance) of innate and adaptive cellular responses were in good agreement with both the favorable safety profile and humoral immunogenicity data (see below). Follow-up studies with the same vaccination regimen as described above were performed in children as young as 2 years as well as older children, adolescents and young adults (up to 45 years) in both non-endemic (Mexico City) and dengue endemic areas (Philippines).33,34 Findings from these studies are consistent with those in the study described above conducted among US adults32 and further show that: flavivirus preimmunity (against Yellow-fever 17D vaccine in Mexico or Japanese encephalitis and dengue in the Philippines) has a positive impact on dengue vaccine immunogenicity without any negative effects on safety; three TDV vaccinations with a 0–3.5–12 month schedule induced robust antibody responses against all four serotypes; and that levels of vaccine virus viremia are consistently low (usually lower than lower limit of quantitation of the PCR or plaque assays used) in all populations. In another third study in Australian adults prior immunity against either dengue 1 or 2 serotypes (conferred by vaccination with conventionally attenuated dengue vaccine candidates developed by Mahidol University, Bangkok, Thailand) primed for a strong and broad response to TDV vaccination against all four serotypes.35 It was also important to address the potential risk that a circulating virus escapes vaccine-induced immunity. The capacity for sera raised against the CYD vaccine viruses to cross-neutralize a large panel of wt strains from each of the four dengue serotypes, collected recently from different areas of dengue endemicity has also been investigated, first with monkey sera.36 Results obtained point to broad coverage of vaccine-induced antibodies against

702

geographically diverse strains, and similar analyses are ongoing with human sera. So far the CYD vaccine candidates have demonstrated a satisfactory safety profile. No serious adverse events (SAE) related to vaccination have been identified in the studies mentioned above. Unlike some whole virion live-attenuated dengue vaccine candidates, the CYD vaccine does not appear to induce the mild dengue-like syndrome associated with viremia.4 Biological safety results have been unremarkable and reactogenicity has appeared similar to that of the control vaccines. Reactogenicity was not increased by the presence of baseline immunity to either dengue or yellow fever, nor was it increased after the second or third vaccination than after the first. While protection against dengue is associated with both humoral and cellular responses, these responses have also been implicated in the immunopathology of severe dengue disease (reviewed in ref. 37). On the antibody side, it was hypothesized that one of the mechanisms responsible for severe disease is the enhancement of viral replication by heterotypic, non-neutralizing antibodies from a prior infection (via the Fc receptor on mononuclear leukocytes—the antibody dependent enhancement (ADE) phenomenon38,39). This point will be specifically addressed below. On the cellular side, immune responses should include high-avidity homologous multivalent responses against all serotypes, should be Th1-biased and dominated by IFNγ over TNFα. In particular, it has been shown that heterologous, cross-reactive responses tend to trigger TNFα while homologous responses trigger IFNγ37,40,41 CD4 and CD8 immune response against the parental YF 17D and dengue viruses elicited by CYD1-4 vaccination were thus assessed in volunteers with and without pre-existing flavivirus immunity.42 The TDV triggers no detectable changes in serum pro-inflammatory cytokines, regardless of the baseline immune status, but induced significant YFV-17D NS3-specific CD8 responses and dengue virus serotype-specific T helper 1 responses, dominated by IFNγ over TNFα. As for antibodies, responses were initially dominated by serotype 4 in baseline-naïve individuals, but subsequent vaccinations broadened the serotype-specific

Human Vaccines Volume 6 Issue 9

responses. A similarly broad response was seen after primary TDV vaccination in participants with preexisting dengue serotype 1 or 2 immunity. Data also demonstrated an absence of cross-reactivity between YF 17D and dengue NS3-specific CD8 responses, and enabled us to identify three new CD8 epitopes in the YF 17D NS3 antigen. According to the original antigenic sin hypotheses, suboptimal, heterologous, anti-NS3 CD8 responses may be involved in the severity of secondary heterologous infection.43 We observed an almost complete absence of CD8 crossreactivity between YF 17D and dengue NS3 antigens, meaning that no potentially deleterious cross-serotype anti-NS responses would be induced by YF17D NS3-expressing CYDs. Subsequent natural infection would therefore boost dengue-specific immune responses with a non-deleterious profile.37,40,41 From a practical point of view, one can mention that the analyses in these studies required a significant amount of blood (from 35 to 50 ml) and would not be applicable to infants or children. We are currently exploring the possibility of performing analyses on a limited (up to 3 mL) amount of blood, in order to be able to analyze responses in children. Phase II and III clinical evaluation. Clinical phase II trials are being conducted on several continents to assessing the safety and immunogenicity of the TDV in children, adolescents and adults with a diverse flavivirus infection and vaccination history and to investigate co-administration with another live virus vaccine—measles, mumps, rubella vaccine—in toddlers. A proof of concept efficacy trial is also underway as part of phase IIb. Four thousand Thai children aged 4–11 years will receive 3 subcutaneous injections of either TDV or control vaccine at 0–6–12 months and will be followed to assess efficacy against virologically-confirmed dengue disease, regardless of the severity. This will be followed by large phase III trials in children and adolescents in Asia and Latin America. In compliance with the WHO guidelines, at each phase of clinical development many subjects will be followed up long term (3–5 years). This will allow us to monitor long term safety including potential severe Dengue, and to assess antibody persistence and the potential need for a booster dose. Antibody Dependent Enhancement The etiology of DHF appears to be multi-factorial. Whether or not antibody dependent enhancement (ADE, see above) is one of these factors in vivo is still a matter of debate (reviewed in ref. 8). We nevertheless took these potential concerns into account and developed early in our DENV vaccine research a sensitive and reproducible in vitro assay using FcγRII positive-K562 cells and flow cytometry. Sera from Thai children vaccinated with firstgeneration LAV candidates were analyzed using this assay and correlated with a low/absent risk linked to ADE activity in vitro, despite very diverse immune profiles, from low to high PRNT levels against one or several dengue serotypes.44 Specifically, in vitro ADE was absent in the presence of broad neutralizing response against all four DENV serotypes. Thus, whatever the role of ADE in the etiology of severe dengue in vivo, a vaccine

www.landesbioscience.com

able to induce a neutralizing response against all four serotypes should circumvent the issue. As stated by WHO guidelines, consensus exists that clinical development of dengue vaccines should not be forestalled by certain hypothetical safety concerns.45 The next round of large efficacy trials with a sample size of several thousand (Phase 3) will provide further evidence of any safety issues occurring in vaccinees versus controls. However, it is likely that this theoretical risk will only be evaluable by Phase 4 trials and post-marketing surveillance, due to the very rare occurrence of such severe outcome. Future Challenges In addition to the direct challenges of clinical development, and as the sanofi pasteur TDV enters clinical phase III, it is important to set appropriate goals. Different additional challenges need to be met to ensure the vaccine’s successful licensure and implementation in the field. Immunology and non-clinical research challenges. We need to better understand the protective mechanisms induced by dengue vaccine candidates. Future clinical efficacy trial data can be used to retrospectively benchmark the preclinical models discussed above, as well as potentially to derive a correlate of protection which would be of great interest. Bridging between NHP data and clinical protection, might facilitate the future development of new vaccine formulations or technologies by reducing the need for further expensive and lengthy clinical trials. With this in mind, a clear immunologically-driven strategy to develop robust, high-throughput, standardized assays is warranted to aid clinical development. Firstly there is a need for high-throughput serotype specific neutralization assays as well as standardized, reproducible cellmediated immunity assays that can be used with small volumes blood, and thus be suitable for application in pediatric studies (see above). A large array of questions related to the evaluation of cellular responses was discussed during a WHO task force on dengue cellular immunity in Bangkok in 2007, generating general recommendations.46 It was stated in particular that, although not mandatory for registration, documenting such responses in clinical trials, including phase 3 trials, would be recommended and would bring important information relative to the shortterm and long-term safety and immunogenicity of the vaccine candidates. Secondly, reliable and simple assays to examine antibodybinding affinity and kinetics might provide useful estimates of the overall avidity of sera against viral or E protein preparations at the polyclonal level. This approach has already been investigated by some investigators, and it would be of interest to compare overall avidity of responses induced by vaccination versus that induced by infection, as well as to compare overall avidity with neutralization antibody results. Isotyping of E-specific IgG subclasses might also be of interest, as factors such as their complement fixing activity could be of importance. Thirdly understanding interference mechanisms between dengue viruses, might lead to simpler vaccination regimens compared with the currently considered regimen, involving multiple

Human Vaccines

703

doses tetravalent vaccine given over 6 to 12 months. Applied preclinical research is therefore needed to (1) better understand the mechanisms of interference involving innate and/or adaptive immunity in suitable in vitro and in vivo models (e.g., through the assay developments mentioned above), (2) use this knowledge to develop new regimens requiring fewer doses over shorter periods and (3=) develop new vaccine technology strategies to alleviate interferences without compromising immunogenicity and safety. Fourthly, non-clinical safety models need to be further explored. While satisfactory mouse and NHP models exist for neurotropism, a similar model for viscerotropism is lacking. Such models could either mimic the immune profile elicited by wt dengue viruses and allow benchmarking of (non-human) safety, biodistribution and protective activity of dengue vaccine candidates or alternatively serve as true pathogenic models for dengue disease, although this latter goal may not be achievable. Institutional and resource challenges. Until recently, interest and investment in dengue disease and vaccination was significantly less than for the ‘big three’ infectious diseases (HIV/ AIDS, malaria, tuberculosis). However, given the increasing global importance of dengue, the globalization of information and, perhaps, the advocacy on global warming and its potential effect on the geographic distribution of diseases, interest in dengue has increased. Advocacy and incentives for are still needed for scientists in academic institutions to work on dengue, and to improve career opportunities in the field. This could be achieved through dedicated funds and grants available for applied research on dengue. Such funds could be provided by public health driven governmental and/or non-governmental organizations, or privately driven foundations such as the Pediatric Dengue Vaccine Initiative (PDVI) which has dedicated dengue vaccine evaluation and access programs. Dengue research has not yet reached a critical mass in many countries where the disease is endemic and ranks among the top five public health concerns. The PDVI, in collaboration with the Pan American Health Organization (PAHO) and WHO, recently organized regional dengue research meetings assembling scientists from Asia and the Americas. Such meetings help gather References Halstead SB. Dengue. Lancet 2007; 370:1644-52. Reiter P. Yellow fever and dengue: a threat to Europe? Euro Surveill 2010; 15:19509. 3. Raviprakash K, Defang G, Burgess T, Porter K. Human Vaccines 2009; 5:520-8. 4. Whitehead SS, Blaney JE, Durbin AP, Murphy BR. Prospects for a dengue virus vaccine. Nat Rev Microbiol 2007; 5:518-28. 5. Guy B, Almond JW. Towards a dengue vaccine: Progress to date and remaining challenges. Comp Immunol Microbiol Infect Dis 2008; 31:239-52. 6. Chambers TJ, Nestorowicz A, Mason PW, Rice CM. Yellow fever/Japanese encephalitis chimeric viruses: construction and biological properties. J Virol 1999; 73:3095-101. 7. Guirakhoo F, Weltzin R, Chambers TJ, Zhang ZX, Soike K, Ratterree M, et al. Recombinant chimeric yellow fever-dengue type 2 virus is immunogenic and protective in nonhuman primates. J Virol 2000; 74:5477-85.

8.

1. 2.

704

9.

10.

11.

12.

data, identify new directions and establish useful synergies and networks to allow dengue scientists and vaccinologists to make significant advances. Efforts such as this need to be renewed, in order to help establish regional scientific networks for dengue-endemic countries. Furthermore, stronger global alliances between the WHO, PAHO, national regulatory authorities, key opinion leaders, leading scientists, private foundations and vaccine developers are needed to share data, regulatory, risk management and licensure requirements in a new partnership model rather the usual private versus public paradigm. Conclusions The Sanofi Pasteur tetravalent live attenuated chimeric virus dengue vaccine candidate has demonstrated satisfactory safety and immunogenicity in both in vitro and in vivo pre-clinical tests, as well as in clinical trials in both flavivirus-naïve and immune individuals. Potential risks, however unlikely, hypothesized as being associated with these chimeric viruses have been explored in depth. Both humoral and cellular responses are induced in humans against all four serotypes and long-term follow-up will address the duration of immunity and theoretical long-term safety issues. An effective vaccine is now urgently required against dengue, and with the initiation of large scale efficacy trials, the present vaccine candidate provides hope that protection is now within our reach. Acknowledgements

The authors would like to acknowledge Grenville Marsh for editorial assistance, all sanofi pasteur and former Acambis team members, as well as external collaborators involved in preclinical, clinical and industrial development, in particular Farshad Guirakhoo, Tom Monath, Steve Higgs, Sutee Yoksan, Veronique Barban, Anke Harenberg, Remi Forrat, Denis Crevat, Gustavo Dayan, Anh Wartel-Tram, Betzana Zambrano, Enrique Rivas and Rafaele Dumas. We would also like to acknowledge all investigators, in particular Drs. Dennis Morrison, Maria Rosario Capeding, Jorge Luis Poo, J Qiao and D Shaw, as well as the volunteers involved in the clinical trials.

Guy B, Guirakhoo F, Barban V, Higgs S, Monath TP, Lang J. Preclinical and clinical development of YFV 17D-based chimeric vaccines against dengue, West Nile and Japanese encephalitis viruses. Vaccine 2010; 28:632-49. Guirakhoo F, Pugachev K, Zhang Z, Myers G, Levenbook I, Draper K, et al. Safety and efficacy of chimeric yellow fever-dengue virus tetravalent vaccine formulations in nonhuman primates. J Virol 2004; 8:4761-75. Barban V, Girerd Y, Aguirre M, Gulia S, Pétiard F, Riou P, et al. High stability of yellow fever 17D-204 vaccine: a 12-year restrospective analysis of large-scale production. Vaccine 2007; 25:2941-50. Pugachev KV, Guirakhoo F, Ocran SW, Mitchell F, Parsons M, Penal C, et al. High fidelity of yellow fever virus RNA polymerase. J Virol 2004; 78:1032-8. Miller BR, Atkins D. Biological characterization of plaque-size variants of yellow fever virus in mosquitoes and mice. Acta Virol 1988; 32:227-34.

13. Vlaycheva LA, Chambers TJ. Neuroblastma celladapted yellow fever 17D virus: Characterization of a viral variant associated with persistent infection and decreased virus spread. J Virol 2002; 76:6172-84. 14. Libraty DH, Pichyangkul S, Ajariyakhajorn C, Endy TP, Ennis FA. Human dendritic cells are activated by dengue virus infection: enhancement by gamma interferon and implications for disease pathogenesis. J Virol 2001; 75:3501-8. 15. Palmer DR, Sun P, Celluzzi C, Bisbing J, Pang S, Sun W, et al. Differential effects of dengue virus on infected and bystander dendritic cells. J Virol 2005; 79:2432-9. 16. Brandler S, Brown N, Ermak TH, Mitchell F, Parsons M, Zhang Z, et al. Replication of chimeric yellow fever virus-dengue serotype 1–4 virus vaccine strains in dendritic and hepatic cells. Am J Trop Med Hyg 2005; 72:74-81. 17. Deauvieau F, Sanchez V, Balas C, Kennel A, de Montfort A, Lang J, et al. Innate immune responses in human dendritic cells upon infection by chimeric Yellow fever Dengue vaccines serotype 1 to 4. Am J Trop Med Hyg 2007; 76:144-54.

Human Vaccines Volume 6 Issue 9

18. Monath TP, Myers GA, Beck RA, Knauber M, Scappaticci K, Pullano T, et al. Safety testing for neurovirulence of novel live, attenuated flavivirus vaccines: infant mice provide an accurate surrogate for the test in monkeys. Biologicals 2005; 33:131-44. 19. Vaughn DW, Green S, Kalayanarooj S, Innis BL, Nimmannitya S, Suntayakorn S, et al. Dengue viremia titer, antibody response pattern and virus serotype correlate with disease severity. J Infect Dis 2000; 181:2-9. 20. Eckels KH, Dubois DR, Putnak R, Vaughn DW, Innis BL, Henchal EA, et al. Modification of dengue virus strains by passage in primary dog kidney cells: preparation of candidate vaccines and immunization of monkeys. Am J Trop Med Hyg 2003; 69:12-6; Erratum in: Am J Trop Med Hyg 2004; 70:336. 21. Putnak RJ, Coller BA, Voss G, Vaughn DW, Clements D, Peters I, et al. An evaluation of dengue type-2 inactivated, recombinant subunit and live-attenuated vaccine candidates in the rhesus macaque model. Vaccine 2005; 23:4442-52. 22. Guy B, Barban V, Mantel M, Aguirre M, et al. Evaluation of interferences between dengue vaccine serotypes in a monkey model. Am J Trop Med Hyg 2009; 80:302-11. 23. Higgs S, Vanlandingham DL, Klingler KA, McElroy KL, McGee CE, Harrington L, et al. Growth characteristics of ChimeriVax-Den vaccine viruses in Aedes aegypti and Aedes albopictus from Thailand. Am J Trop Med Hyg 2006; 75:986-93. 24. Seligman SJ, Gould EA. Live flavivirus vaccines: reasons for caution. Lancet 2004; 363:2073-5. 25. Monath TP, Kanesa-Thasan N, Guirakhoo F, Pugachev K, Almond J, Lang J, et al. Recombination and flavivirus vaccines: a commentary. Vaccine 2005; 23:2956-8. 26. Hombach J, Kurane I, Wood D. Arguments for live flavivirus vaccines. Lancet 2004; 364:498-9. 27. Taucher C, Berger A, Mandl CW. A trans-complementing recombination trap demonstrates a low propensity of flaviviruses for intermolecular recombination. J Virol 2010; 84:599-611. 28. McGee CE, Tsetsarkin K, Vanlandingham DL, McElroy KL, Lang J, Guy B, et al. Substitution of wild-type yellow fever Asibi sequences for 17D vaccine sequences in ChimeriVax-dengue 4 does not enhance infection of Aedes aegypti mosquitoes. J Infect Dis 2008; 197:686-92.

www.landesbioscience.com

29. McGee CE, Lewis MG, Claire MS, Wagner W, Lang J, Guy B, et al. Recombinant chimeric virus with wildtype dengue 4 virus premembrane and envelope and virulent yellow fever virus Asibi backbone sequences is dramatically attenuated in nonhuman primates. J Infect Dis 2008; 197:693-7. 30. Lindsey NP, Schroeder BA, Miller ER, Braun MM, Hinckley AF, Marano N, et al. Adverse event reports following yellow fever vaccination. Vaccine 2008; 26:6077-82. 31. Guirakhoo F, Kitchener S, Morrison D, Forrat R, McCarthy K, Nichols R, et al. Live attenuated chimeric yellow fever dengue type 2 (ChimeriVax-DEN2) vaccine: Phase I clinical trial for safety and immunogenicity: effect of yellow fever pre-immunity in induction of cross neutralizing antibody responses to all 4 dengue serotypes. Hum Vaccine 2006; 2:60-7. 32. Morrison D, Legg TJ, Billings CW, Forrat R, Yoksan S, Lang J. A novel tetravalent dengue vaccine is well tolerated and immunogenic against all 4 serotypes in flavivirus-naive adults. J Infect Dis 2010; 201:370-7. 33. Crevat D, Reynolds D, Langevin E, Capeding MR. Safety and Immunogenicity of a tetravalent dengue vaccine in flavivirus-naive and immune pediatric populations with two vaccination regimens. Am J Trop Med Hyg 2009; 81:113. 34. Forrat R, Poo JL, Galán Herrera JF. Immune response to tetravalent dengue vaccination in Mexican subjects: The effects of yellow fever vaccination. Am J Trop Med Hyg 2008; 79:114. 35. Qiao J, Shaw D, Forrat R, Wartel-Tram A, Lang J. Safety, viremia and immunogenicity of sp’s TDV in adults previously exposed to live, attenuated flaviviruses. 13th International Congress on Infectious Diseases 2008; Kuala Lumpur, Malaysia. 36. Barban V, Mantel M, Gulia S, Girerd Y, Boaz M, Crevat D, et al. Evaluation of neutralizing antibody responses against a large range of wild-type isolates in sera of primates vaccinated with CYD TV DEN vaccine. 58th ASTM&H annual meeting Washington DC 2009. 37. Green S, Rothman A. Immunopathological mechanisms in dengue and dengue hemorrhagic fever. Curr Opin Infect Dis 2006; 19:429-36.

Human Vaccines

38. Halstead SB, O’Rourke EJ. Antibody-enhanced dengue virus infection in primate leukocytes. Nature 1977; 265:739-41. 39. Kliks SC, Nisalak A, Brandt WE, Wahl L, Burke D. Antibody-dependent enhancement of dengue virus growth in human monocytes as a risk factor for dengue hemorrhagic fever. Am J Trop Med Hyg 1989; 40:44451. 40. Mangada MM, Endy TP, Nisalak A, Chunsuttiwat S, Vaughn DW, Libraty DH, et al. Dengue-specific T cell responses in peripheral blood mononuclear cells obtained prior to secondary dengue virus infections in Thai schoolchildren. J Infect Dis 2002; 185:1697-703. 41. Mangada MM, Rothman AL. Altered cytokine responses of dengue-specific CD4+ T cells to heterologous serotypes. J Immunol 2005; 175:2676-83. 42. Guy B, Nougarede N, Begue S, Sanchez V, Souag N, Carre M, et al. Cell-mediated immunity induced by chimeric tetravalent dengue vaccine in naive or flavivirus-primed subjects. Vaccine 2008; 26:5712-21. 43. Mongkolsapaya J, Dejnirattisai W, Xu XN, Vasanawathana S, Tangthawornchaikul N, Chairunsri A, et al. Original antigenic sin and apoptosis in the pathogenesis of dengue hemorrhagic fever. Nat Med 2003; 9:921-7. 44. Guy B, Chanthavanich P, Gimenez S, Sirivichayakul C, Sabchareon A, Begue S, et al. Evaluation by flow cytometry of antibody-dependent enhancement (ADE) of dengue infection by sera from Thai children immunized with a live-attenuated tetravalent dengue vaccine. Vaccine 2004; 22:3563-74. 45. Edelman R, Hombach J. Guidelines for the clinical evaluation of dengue vaccines in endemic areas. Summary of a WHO technical consultation. Vaccine 2008; 26:4113-9. 46. Thomas SJ, Hombach J, Barrett A. Scientific consultation on cell mediated immunity (CMI) in dengue and dengue vaccine development. Vaccine 2009; 27:35568.

705