afety and immunogenicity of a tetravalent dengue vaccine in healthy hildren aged 2 – 11 years in Malaysia : A randomized , lacebo-controlled , Phase III study

Background: Dengue disease is a major public health problem across the Asia-Pacific region for which there is no licensed vaccine or treatment. We evaluated the safety and immunogenicity of Phase III lots of a candidate vaccine (CYD-TDV) in children in Malaysia. Methods: In this observer-blind, placebo-controlled, Phase III study, children aged 2–11 years were randomized (4:1) to receive CYD-TDV or placebo at 0, 6 and 12 months. Primary endpoints included assessment of reactogenicity following each dose, adverse events (AEs) and serious AEs (SAEs) reported throughout the study, and immunogenicity expressed as geometric mean titres (GMTs) and distribution of dengue virus (DENV) neutralizing antibody titres. Results: 250 participants enrolled in the study (CYD-TDV: n = 199; placebo: n = 51). There was a trend for reactogenicity to be higher with CYD-TDV than with placebo post-dose 1 (75.4% versus 68.6%) and post-dose 2 (71.6% versus 62.0%) and slightly lower post-dose 3 (57.9% versus 64.0%). Unsolicited AEs declined in frequency with each subsequent dose and were similar overall between groups (CYD-TDV: 53.8%; placebo: 49.0%). Most AEs were of Grade 1 intensity and were transient. SAEs were reported by 5.5% and 11.8% of participants in the CYD-TDV and placebo groups, respectively. No deaths were reported. Baseline seropositivity against each of the four DENV serotypes was similar between groups, ranging from 24.0% (DENV-4) to 36.7% (DENV-3). In the CYD-TDV group, GMTs increased post-dose 2 for all serotypes compared with baseline, ranging from 4.8 (DENV-1) to 8.1-fold (DENV-3). GMTs further increased postdose 3 for DENV-1 and DENV-2. Compared with baseline, individual titre increases ranged from 6.1-fold (DENV-1) to 7.96-fold (DENV-3). Conclusions: This study demonstrated a satisfactory safety profile and a balanced humoral immune response against all four DENV serotypes for CYD-TDV administered via a three-dose regimen to children in Malaysia. This is an open-access article distributed under the terms of the Creative ommons Attribution License, which permits unrestricted use, distribution, and eproduction in any medium, provided the original author and source are credited. Trial registration: The trial is registered on ClinicalTrials.gov.: National Clinical rials Identifier (NCT ID): NCT01254422. ∗ Corresponding author. Tel.: +65 6225 3836/6431 2396; fax: +65 6431 2445. E-mail addresses: amarhss@gmail.com (A.-S. HSS), ohmt@um.edu.my (M.-T. Koh), kktalk@gmail.com (K.K. Tan), r chanlg@hotmail.com (L.G. Chan), Linghua.Zhou@sanofipasteur.com 264-410X/$ – see front matter © 2013 The Authors. Published by Elsevier Ltd. All rights ttp://dx.doi.org/10.1016/j.vaccine.2013.10.013 © 2013 The Authors. Published by Elsevier Ltd. All rights reserved.


Introduction
Dengue disease is caused by dengue virus (DENV) serotypes (1)(2)(3)(4) and is transmitted mainly by Aedes aegypti mosquitoes [1]. Dengue disease is classified as with or without warning signs, or severe dengue disease [2]. Infection with one DENV serotype usually imparts immunity to that serotype [3]. No licensed vaccine or treatment for dengue disease currently exists. Prevention relies on individual protection and vector control measures, which have limited effectiveness [1].
The primary objectives of our study were to evaluate the safety and immunogenicity of CYD-TDV in children aged 2-11 years in Malaysia. This is the first Phase III CYD-TDV study to be reported that uses Phase III lots (manufactured by large-scale production processes) in an area where dengue disease is endemic. Additionally, one of the study sites was located in the Sarawak state of Malaysia, where Japanese Encephalitis (JE) is endemic and JE vaccination is routine.

Study design and participants
This multicentre, randomized, observer-blind, placebocontrolled, Phase III study was conducted between December 2010 and August 2012 at four sites in Malaysia: Kuala Lumpur, Ipoh (Perak state), Seremban (Negeri Sembilan state) and Kuching (Sarawak state). The methodology was similar to that of previous CYD-TDV Phase II studies [22,26] but used Phase III vaccine lots. Healthy children (aged 2-11 years) were assigned randomly to two groups (4 CYD-TDV: 1 placebo). All participants received three injections at 0, 6 and 12 months and were followed up for 6 months post-dose 3 to assess safety.
Participants were recruited by the study investigators, subinvestigators and research nurses. Randomization was performed by study-site personnel via an Interactive Voice Recognition System (IVRS), using a permuted block method with stratification by centre and age. A double randomization system separated treatment allocation from doses dispensed. The IVRS was used in a blind manner to ensure a balanced distribution of the number of participants with a history of JE vaccination, known JE infection or dengue disease infection in each age group (2-5 and 6-11 years) of the study groups, based on medical records or parents' recollection.
Girls of childbearing age were checked for pregnancy (urine test) and required to abstain from sexual intercourse or use contraception from four weeks before the first dose until four weeks after the last dose. Exclusion criteria included prior or current participation in another clinical study; receipt of blood or blood-derived products in the previous 3 months that could interfere with immunogenicity assessments; hypersensitivity to the vaccine components; and vaccination with any other vaccine (except for pandemic influenza) in the 4 weeks prior to enrolment. Contraindications to receiving subsequent vaccinations included significant allergic reaction (AR), serious adverse event (SAE) or ongoing adverse event (AE) related to the previous vaccination.

Ethical considerations
The study was conducted in accordance with the Declaration of Helsinki (Seoul revision), with Good Clinical Practice (defined by the International Conference on Harmonisation), and with applicable national and local requirements. The Medical Research and Ethics Committee of the Ministry of Health of Malaysia approved the protocol. Each investigator obtained approval from an independent ethics committee or institutional review board. Participants' parents or legal representatives provided written informed consent and children aged ≥7 years signed an assent form. The trial is registered on ClinicalTrials.gov. (NCT01254422).

Study procedures
Blood samples were taken prior to vaccination, to assess flavivirus (FV) serostatus (dengue disease or JE) using plaque reduction neutralization tests (PRNT 50 ) [30,31], and 28 days post-dose 2 and 3 to assess immunogenicity.

Safety and reactogenicity
Safety and reactogenicity data (graded in severity from 1 to 3) were collected after each dose: immediate AEs within 30 min; solicited injection-site reactions within 7 days; solicited systemic reactions within 14 days; and unsolicited AEs within 28 days of vaccination. SAEs and suspected dengue disease cases requiring hospitalization were monitored throughout the study. An independent data monitoring committee reviewed all SAEs.

Immunogenicity
Neutralizing antibody levels against DENV1-4 were assessed 28 days post-dose 2 and 3 using a PRNT 50 compliant with WHO guidelines [30,31] and expressed as geometric mean titres (GMTs) and seropositivity rates (percentage of participants with titres ≥10 1/dil). Distribution of individual titres at enrolment and postdose 2 and 3 of CYD-TDV or placebo was analyzed using reverse cumulative distribution curves.

Secondary endpoints
CYD-TDV safety and immunogenicity assessments were stratified according to baseline FV-serostatus (seropositive or seronegative for dengue disease or JE) and by age (2-5 and 6-11 years). Pre-vaccination, post-dose 2 and post-dose 3 GMTs were also analyzed according to FV-serostatus at baseline (seropositive or seronegative for dengue disease and JE).

Sample size and study populations
With a planned sample size of 250 participants (CYD-TDV: n = 200; placebo: n = 50) and assuming a drop-out rate of 15%, the probability of observing a common AE after three vaccinations was 100% and 82% if the incidence of the AE was 5% and 1%, respectively. Of the 200 CYD-TDV recipients, 60-140 were expected to be FV-seropositive at baseline. Based on a variability of 0.7 log 10 for the PRNT 50 assay, the power to detect differences of ≥0.4 in log 10 GMTs between FV-seropositive and -seronegative participants and between age groups was >90%. The populations analyzed included a Safety Analysis Set (SAS, participants receiving at least one dose of study vaccine) and a Full Analysis Set (FAS, participants receiving at least one dose of study vaccine, with a valid post-vaccination serology result). Further analyses were conducted stratified by FV-serostatus and by age group.

Statistical analyses
Analyses were descriptive with no hypothesis tested. For the main parameters, 95% confidence intervals of point estimates were calculated using a normal approximation for quantitative data and exact binomial distribution (Clopper-Pearson method) for proportions [32,33]. Analyses were conducted with SAS software, version 9.1 or above (SAS Institute, Cary, NC, USA).

Results
The first participant was enrolled on 2 December 2010 and the last participant's 6 month follow-up visit was 14 August 2012. We enrolled 250 participants in the study (CYD-TDV: n = 199; placebo: n = 51), of whom 196 (98.5%) and 50 (98%) in the CYD-TDV and placebo group, respectively, completed the vaccination phase and were included in the FAS (n = 246, Fig. 1).
In the SAS (n = 250), 55.8% of participants in the CYD-TDV group and 60.8% in the placebo group were FV-seropositive (seropositive for DENV and/or JE). In the CYD-TDV and placebo groups, 49.7% and 52.0% of participants were aged 2-5 years, respectively. In the CYD-TDV group, 57.3% of children aged 2-5 years were FV-seropositive and 42.7% were FV-seronegative and in children aged 6-11 years, 54.0% were FV-seropositive and 46.0% were FV-seronegative.

Enrolled: n=250
Randomized to CYD-TDV group n=199 Randomized to placebo group n=51 Received CYD-TDV at visit 01: n=199 (100%)  . 1. Study flow chart: progress of participants through the study. Reasons for withdrawal: In the CYD-TDV group, one participant withdrew due to an AE (allergic conjunctivitis), one was lost to follow-up after the first vaccination, and one did not comply with the protocol by refusal of a blood sample before injection 3. In the placebo group, one participant experienced a SAE (right VII nerve paralysis) after the first vaccination, which was assessed by the investigator to be related to treatment. Baseline demographic characteristics were similar between groups and in the FV-serostatus subsets (Table 1). However, baseline GMTs against DENV1-4 were slightly higher in those aged 6-11 years than in those aged 2-5 years ( Table 2). The majority of participants were of Asian origin (n = 248) and two participants were of Asian/Caucasian origin.

Safety and reactogenicity
The proportion of participants reporting solicited reactions (total and injection site and systemic reactions) was similar in the CYD-TDV (89.4%) and placebo (94.1%) groups (Table 3). However, there was a trend for slightly higher reactogenicity with CYD-TDV compared with placebo post-dose 1 and 2 (75.4% versus 68.6% and 71.6% versus 62.0%, respectively) and slightly lower reactogenicity in the CYD-TDV group compared with placebo post-dose 3 (57.9% versus 64.0%, Fig. 2).
The number of solicited systemic reactions decreased after each injection in both groups. Solicited injection site reactions in the CYD-TDV group tended to be more frequent post-dose 2 compared with post-dose 1 and less frequent post-dose 3. In the placebo group, solicited injection site reactions were more frequent postdose 2 and 3 than post-dose 1 (Fig. 2).
Most unsolicited AEs and ARs were non-serious and did not lead to study discontinuation (Table 3). SAEs were reported by 5.5% of participants in the CYD-TDV group and 11.8% in the placebo group (Table 3). All SAEs were assessed as unrelated to study vaccine, except for an SAE of VII nerve paralysis in the placebo group. This participant did not receive any further vaccinations and recovered 4 months later. No deaths were reported.

Safety and reactogenicity by baseline FV-serostatus and age
An analysis by subset showed that safety and reactogenicity data were not markedly affected by FV-serostatus or age (Fig. 2). The decrease in the incidence of systemic reactions post-dose 3 appeared more marked in the FV-seropositive group than the FVseronegative group.
Robust immune responses were reported after vaccination with CYD-TDV regardless of FV-serostatus ( Table 2). The geometric mean fold rise of individual antibody titres ranged from 4.8-fold (DENV-1) to 8.1-fold (DENV-3) post-dose 2 and from 6.1-fold (DENV-1) to 8.0-fold (DENV-3) post-dose 3. Post-dose 2 and 3 GMTs were higher in the FV-seropositive group. Post-dose 3 GMTs showed a greater increase in the FV-seronegative group than in the FV-seropositive group. Pre-vaccination and post-dose 3 GMTs were higher in participants who were seropositive for dengue disease at baseline than in participants who were seronegative, regardless of JE serostatus. Post-dose 3 GMTs were slightly higher in JE seropositive participants versus JE seronegative participants. However, this difference was not significant. The highest post-dose 3 GMTs were observed in participants who were seropositive for both dengue disease and JE at baseline. Robust immune responses were also reported regardless of age, although post-dose 2 and 3 GMTs were higher in [6][7][8][9][10][11] year-olds compared with 2-5 year-olds ( Table 2).
The post dose 2 and 3 antibody reverse cumulative distribution curves demonstrated good similarity of responses (curves close together and parallel, Fig. 3).In FV-seronegative children, the dose 3 vaccination increased the overall neutralizing titres with a left shift of the curves for serotypes 1 and 2.

Discussion
CYD-TDV is in advanced stages of clinical development and large Phase III efficacy studies are underway in Latin America (NCT01374516) and Asia (NCT01373281); no other dengue disease candidate vaccine has reached this stage of development to date. As the first paediatric trial to use Phase III CYD-TDV lots manufactured using large-scale processes, the present study is a key component of the global clinical trial programme for CYD-TDV [17][18][19]. Results showed that CYD-TDV had a satisfactory safety profile and generated a balanced humoral immune response against DENV1-4 in this paediatric population.
All four DENV serotypes circulate in Malaysia and their relative geographic distributions vary over time [11]. Baseline seropositivity against DENV1-4 was similar in both groups, which was expected, given that dengue disease is endemic in Malaysia.
The safety profile of CYD-TDV was satisfactory for the total population and in the subsets analyzed. Reactogenicity decreased with subsequent doses of CYD-TDV, consistent with observations from previous phase I and II studies [11]. Injection site erythema and swelling were reported frequently, but at Grade 1 intensity and at similar rates in both treatment groups. Reactogenicity was similar in the subsets of FV-serostatus and age ranges and in the whole population, without increased reactogenicity in 2-5 year-olds, or baseline FV-seropositive children compared with 6-11-year-olds or FV-seronegative children. The favourable safety and reactogenicity results observed are similar to those from previous Phase I and II studies [20][21][22][23][24][25][26][27][28][29]. Our data from Malaysia are consistent with those reported from 2 to 11-year olds in Peru [22] and Singapore [26], from 4 to 11-year olds in Thailand [29] and from 9 to 16-year olds in Latin America [28]. The three-dose regimen of CYD-TDV elicited a good immune response in terms of GMTs, regardless of FV-serostatus or age group. GMTs increased both post-dose 2 and 3 in the overall population. The robust and balanced antibody responses against DENV1-4 were similar to those reported in other studies of CYD-TDV in children [22,26,28]. However, the results of the Phase IIb proof-of-concept efficacy study by Sabchereon et al. challenged the hypothesis that such a robust, balanced antibody response profile, as assessed by PRNT 50 , translates to similar levels of protection against all viruses of each DENV serotype [29]. The higher GMTs reported after vaccination in the FV-seropositive group compared with the FV-seronegative group were also observed in children vaccinated in Latin America [28]. The third dose had little impact in the FV-seropositive group but was more marked in the FV-seronegative group. Therefore, in the context of a mixed population with both FV-seropositive and FV-seronegative children, a three-dose regimen is beneficial to induce a balanced immune response. Baseline serostatus for dengue disease but not for JE had an impact on postdose 3 GMTs. However, it should be noted that the sample size of this subset was limited-only 67 participants in the CYD-TDV group and 17 participants in the placebo group were seropositive for JE antibodies (PRNT 50 titre ≥ 10 l/dil). The immune response in the current study was satisfactory in both age groups, although higher GMTs were reported in those aged 6-11-years olds than 2-5-year olds post-dose 2 and 3. This observation was probably the result of the difference in baseline DENV serostatus between the two age groups (difference in pre-vaccination GMTs for DENV1-4) as the main driver for vaccine immunogenicity. The link between baseline FV-serostatus and age range on the immune response is presumed to be the result of a longer period of exposure to natural dengue infection in the 6-11 years group compared with the younger cohort. Although the highest attack rate for dengue infection is commonly observed in young adults aged 20-24 years old [34], the large proportion of children who were seropositive at baseline for at least one DENV serotype indicates that dengue disease is endemic and infects children in Malaysia at a young age. However, no suspected cases of dengue disease requiring hospitalization were reported by the investigators or participants' parents in this study and antibody titres in the placebo group did not increase, suggesting the absence or very limited circulation of wild-type DENV strains in this population during the study period. Therefore, the results of vaccine immunogenicity were probably not biased by natural dengue infections. However, the placebo group was small and conclusions based on these results need to be kept in context. This study was not designed to assess CYD-TDV efficacy and did not address long-term vaccine safety or immune persistence, with follow-up limited to 6 months. However, large Phase III efficacy studies are underway and long-term follow-up is ongoing in these studies, as well as in other Phase I [20] and Phase II studies [29]. Furthermore, the sample size of this study was too small to accurately analyze any potential priming effect of previous exposure to individual FV subtypes (e.g. JE).

Conclusion
In conclusion, this Phase III study demonstrated a satisfactory safety profile and a balanced humoral immune response against DENV1-4 for CYD-TDV administered via a three-dose regimen to children in Malaysia living in an area where dengue disease is endemic.
Global Pharmacovigilance and Epidemiology, Marcy l'Etoile, Lyon, France and Mandy Khoo from Sanofi Pasteur Regional Study Management and Logistic, Sanofi Pasteur Singapore, for co-ordinating study activities. The authors also thank all of the children who participated in the study and their parents/guardians, and the study-site personnel for their contributions.
The authors take full responsibility for the content of this contribution and thank Communigen Ltd. (supported by Sanofi Pasteur) for editorial assistance with the preparation of this manuscript and inputting author comments. The authors also thank Grenville Marsh for providing critical comments and suggestions on the draft. Contributions: The authors made the following contributions to the study: study conception and design (AB, DC, YH, KKT), laboratory data acquisition (MTK), patient enrolment/patient data acquisition (AB, DC, LGC, MTK, YH, ASHSS, KKT) and study data analysis and interpretation (AB, DC, MTK, YH, ASHSS, LZ). All authors contributed to the conceptualization and drafting of the article, participated in the critical review of the article and approved the final version submitted for publication. All authors were involved in the decision to submit to Vaccine. Conflict of interest: Lynn Zhou, Alain Bouckenooghe, Denis Crevat and Yanee Hutagalung are employees of Sanofi Pasteur. Alain Bouckenooghe and Yanee Hutagalung own stock options in Sanofi Pasteur. Mia-Tuang Koh has received financial support from Sanofi Pasteur to travel to meetings. Tan Kah Kee has received financial support from Sanofi Pasteur to travel to meetings and has received study equipment and study drugs. Amar-Singh HSS and Lee Gaik Chan declare no conflicting interests. Role of the funding source: This study was funded by Sanofi Pasteur who also contributed to the study design, data collection, analysis and interpretation, review of the manuscript and decision to publish.