The development of a safe and effective tetravalent dengue vaccine has been a goal for decades, yet a licensed vaccine is still not available. At a minimum, such a vaccine must have an acceptable safety profile and must induce long-lived protective immunity against all four serotypes of wild type dengue virus. Ideally this would be accomplished with a single dose of vaccine. However, tetravalent vaccines currently in clinical development require two or three doses 
. A number of dengue viruses have been attenuated either by serial passage in tissue culture or by the introduction of attenuating mutations and/or chimerization of dengue viruses using recombinant DNA technology 
. In addition, a tetravalent vaccine based on chimerization of DENV-1-4 with the yellow fever vaccine virus is currently in Phase II/III clinical development 
. Other investigational vaccines have been abandoned because they were either under or over-attenuated 
. In addition, there are concerns that viral interference may affect the ability of a live attenuated tetravalent dengue vaccine to induce a balanced immune response to all four serotypes. For these reasons, we have attempted to carefully examine the safety profile and humoral immune response to each monovalent DENV vaccine candidate virus prior to formulating a tetravalent vaccine. Clinical examination and laboratory studies performed every other day for the first 16 days post vaccination have helped to fully characterize the reactogenicity and kinetics of replication of these viruses when given as monovalent vaccines.
The ability of live attenuated dengue vaccines to induce a satisfactory antibody response without clinically significant reactogenicity has been a hurdle to DENV vaccine development. There are few published studies of monovalent DENV-1 vaccines to which we can compare the safety and immunogenicity of rDEN1Δ30 in healthy adult subjects. Three such studies describe the DENV-1 vaccine candidates 16007, 45AZ5 PDK20, and 45AZ5 PDK27 
. These candidates were attenuated by serial passage in tissue culture, and the parent virus of 45AZ5 PDK20 and 45AZ5 PDK27 underwent chemical mutagenesis prior to passage in tissue culture. Five subjects received DENV-1 16007 as a monovalent vaccine and 12 subjects received 45AZ5 PDK20. Vaccine candidate 16007 was well tolerated: none of the subjects developed an oral temperature >38°C, one subject developed a rash and 2 subjects developed elevated liver function tests 
. However, only 60% of vaccinees seroconverted to the vaccine as defined as a PRNT50
10.In contrast, 100% of vaccinees seroconverted to DEN1 after receipt of the monovalent DENV-1 vaccine 45AZ5 PDK20; however, 5/12 subjects developed a temperature >38°C and 3/12 developed dengue-like illness 
. Additionally, 20% to 35% of subjects who received a tetravalent formulation containing 45AZ5 PDK20 developed a temperature of >38°C 
. The further passaged DENV-1 candidate vaccine 45AZ5 PDK27 was evaluated as a monovalent DENV-1 vaccine in 10 subjects and was found to be more attenuated than 45AZ5 PDK20 
. While the reactogenicity profile of 45AZ5 PDK27 was acceptable, only 40% of vaccinees seroconverted to DENV-1 following a single dose of 45AZ5 PDK27 
. Although 45AZ5 PDK27 was less immunogenic than 45AZ5 PDK20, it has replaced 45AZ5 PDK27 in the tetravalent formulations currently under evaluation, illustrating the trade-off that is sometimes necessary to balance reactogenicity and immunogenicity. Thus, achieving a satisfactory balance between attenuation and immunogenicity has been difficult to attain by passage and/or mutagenesis of DENV-1. In the rDEN1Δ30 vaccine virus, the attenuating 30 nucleotide deletion is located in the 3′ UTR but the virus is otherwise wild-type. This might have contributed to its high level of infectivity and immunogenicity.
The promising safety profile of the rDEN1Δ30 vaccine described in a previous single-dose study was affirmed in the current two-dose study 
. None of the subjects developed fever, a dengue-like illness or any study-related SAE. All of the clinical solicited adverse events were mild or moderate in severity and were transient in nature. As we have reported for our previous dengue vaccine studies, neutropenia and rash were the most commonly reported adverse events in subjects who received rDEN1Δ30. As was seen in the previous study of this vaccine candidate, the onset of neutropenia and rash generally followed the onset of viremia. The definition of neutropenia used in our studies is more inclusive than that utilized in many other dengue vaccine studies. We defined neutropenia as an ANC of ≤1500/mm3
and obtained blood counts every other day for the first 16 days after vaccination; other studies of leading dengue vaccine candidates have defined neutropenia as an ANC<1,000/mm3
with less frequent monitoring, e.g. only on day 15, or on days 4, 8, 12, and 16 post vaccination 
. Had we used an ANC<1,000/mm3
to define neutropenia our studies, only 2/71 (ANC measured on day 16 only) or 5/71 (ANC measured on days 4, 8, 12, and 16 only) vaccine recipients would have been described as neutropenic, a rate of neutropenia comparable to that reported in the studies indicated above. In addition, more than 60% of our study subjects were of African descent, a population known to maintain a lower mean baseline ANC 
. In comparison, subjects of African descent made up only 5–33% of volunteers in the studies referenced above. Most importantly, the very short duration of neutropenia observed in recipients of rDEN1Δ30 was not associated with fever or other clinical signs or symptoms suggestive of neutropenia-related illness. In contrast to neutropenia associated with neoplasia or therapeutic interventions, we assume that neutrophil function during DENV infection or following vaccination with a live-attenuated DENV vaccine is not compromised and that the short duration of neutropenia does not compromise protective immune functions. Our monovalent DENV vaccines have been evaluated in approximately 500 subjects thus far and fever of unknown origin or any sign or symptom suggestive of opportunistic infections has not been observed. To our knowledge, neutropenia due to natural dengue illness has not been associated with secondary bacteremia. A review of cases of concomitant bacteremia in patients diagnosed with DHF found that leukopenia was not associated with an increased incidence of bacteremia compared with control patients 
The investigational vaccine rDEN1Δ30 appears to have a well-balanced safety and immunogenicity profile. In addition to the acceptable reactogenicity of rDEN1Δ30 described above, the vaccine was able to induce an overall seroconversion rate of 93% (65/70 subjects) when administered as a single subcutaneous dose to flavivirus-naïve healthy adults. Importantly, the immunogenicity end-point used in both this and our previously reported study of rDEN1Δ30 was a 4-fold rise in serum neutralizing antibody (i.e., a PRNT60
titer of ≥1
20 defined seroconversion) and is more stringent compared with a PRNT50
titer of ≥1
10 used by others in the studies described above. In addition to the high seroconversion rate that was achieved following a single dose of vaccine, rDEN1Δ30 induced sterilizing humoral immunity to infection with a second dose of vaccine administered 4 or 6 months later in the vast majority of vaccinees, a finding that could be indicative of long-term homotypic protection as has been described following natural DENV infection 
In the present study, the four subjects who did not develop detectable neutralizing antibody after the first dose of rDEN1Δ30 could not be infected by a second dose of vaccine given 4 months later. Following dose 1, one of the 4 subjects had laboratory findings consistent with infection by the vaccine virus, and another subject had findings that may be suggestive of infection by the vaccine virus. The former had detectable viremia on study day 12 and a mild neutropenia on study day 16, and the latter subject was mildly neutropenic on study day 4 following first vaccination. The other two subjects did not have clinical or laboratory findings suggestive of infection. Of these four subjects, three received a second dose of vaccine but did not develop a detectable neutralizing antibody response following dose 2. It is unclear whether these individuals were vaccine non-responders, completely resistant to infection, or whether their innate immune system was able to abort infection prior to the induction of an antibody response. Loss of vaccine potency was excluded as a cause for the absence of infection following dose 2.
This study has important implications for the development of a live attenuated tetravalent dengue vaccine. First, the expanded safety evaluation of the monovalent rDEN1Δ30 vaccine confirmed it to be well tolerated without vaccine-related fever or dengue-like illness. Second, the vaccine was able to induce seroconversion to DENV-1 in 93% of recipients when given as a single subcutaneous dose of 103 PFU. The 103 PFU dose required for the induction of seroconversion to DENV-1 and the ability to grow this virus to high titer (>107 PFU/mL), make this a very economical vaccine component to produce. Third, the vaccine virus behaved immunologically much like wild-type virus in that sterilizing homotypic immunity was induced after a single dose. It is unknown whether this will make boosting of the humoral response to DENV-1 more difficult should multiple doses of the tetravalent vaccine be required. It is also not known whether the observed protection against a second dose of vaccine virus will translate into protection against infection/disease following exposure to wild type DENV-1 virus after vaccination. Future studies with tetravalent vaccine will need to evaluate whether boosting is necessary and will need to be designed to determine a suitable interval for boosting. Lastly, because the replication kinetics of this virus have been very well characterized as a monovalent vaccine, any effect of viral interference on the replication kinetics of the vaccine when administered as part of a tetravalent formulation should be discernable. In summary, rDEN1Δ30 is an excellent candidate for inclusion in live tetravalent dengue vaccine formulations, and clinical evaluation of tetravalent dengue vaccines with this rDEN1Δ30 as DENV-1 component have been initiated.
Our study has at least two potential limitations regarding the infectivity, safety, and immunogenicity of rDEN1Δ30 when administered as a component of a tetravalent formulation. First, sterilizing humoral immunity to a second dose of monovalent rDEN1Δ30 vaccine at 6 months might not translate into sterilizing humoral immunity six months after vaccination with a tetravalent vaccine. It is not known whether or not the high seroconversion rates achieved following a single dose of the rDEN1Δ30 vaccine, when given as a monovalent virus, will be achievable when the candidate vaccine is included in a tetravalent formulation. Secondly, the results of our studies in flavivirus-naive healthy adults do not necessarily mirror how this investigational vaccine will behave in flavivirus-naive and partially immune children, one of the target groups for vaccination in several hyperendemic areas. Tetravalent vaccine studies will need to carefully proceed from adults into children and from flavivirus-naive into partially immune individuals to evaluate the safety and immunogenicity of this vaccine in these vulnerable populations.