Combination Measles-Mumps-Rubella-Varicella Vaccine in Healthy Children

Abstract A combined measles-mumps-rubella-varicella (MMRV) vaccine is expected to facilitate universal immunization against these 4 diseases. This study was undertaken to synthesize current research findings of the immunogenicity and safety of MMRV in healthy children. We searched PubMed, Embase, BIOSIS Previews, Web of Science, Cochrane Library, and other databases through September 9, 2014. Eligible randomized controlled trials (RCTs) were selected and collected independently by 2 reviewers. Meta-analysis was conducted using Stata 12.0 and RevMan 5.3. Twenty-four RCTs were included in qualitative synthesis. Nineteen RCTs compared single MMRV dose with measles-mumps-rubella vaccine with or without varicella vaccine (MMR + V/MMR). Similar seroconversion rates of these 4 viruses were found between comparison groups. There were comparable geometric mean titers (GMTs) against mumps and varicella viruses between MMRV group and MMR + V/MMR group. MMRV group achieved enhanced immune response to measles component, with GMT ratio of 1.66 (95% confidence interval [CI] 1.48, 1.86; P < 0.001) for MMRV versus MMR and 1.62 (95% CI 1.51, 1.70; P < 0.001) for MMRV versus MMR + V. Meanwhile, immune response to rubella component in MMRV group was slightly reduced, GMT ratios were 0.81 (95% CI 0.78, 0.85; P < 0.001) and 0.79 (95% CI 0.76, 0.83; P < 0.001), respectively. Well tolerated safety profiles were demonstrated except higher incidence of fever (relative risks 1.12–1.60) and measles/rubella-like rash (relative risks 1.44–1.45) in MMRV groups. MMRV had comparable immunogenicity and overall safety profiles to MMR + V/MMR in healthy children based on current evidence.

The MMRV vaccine is expected to offer several benefits: simplifying immunization delivery; increasing compliance with immunization; decreasing cumulative exposure to additives; and reducing healthcare costs. [5][6][7] Up to date, although several reviews 5,7 -10 focused on the immunogenicity and safety of ProQuad and/or Priorix-Tetra have been published, we lack a systematic understanding of MMRV vaccine. This study was conducted as a meta-analysis of clinical trials to investigate the immunogenicity and safety of 1 and 2-dose vaccination courses of MMRV vaccine in healthy children.

Eligibility Criteria
Eligible study designs were randomized controlled trials (RCTs) comparing MMRV-vaccinated children (MMRV group) with MMR-vaccinated (MMR group) or MMR þ varicella vaccine coadministered children (MMR þV group). The population of interest were healthy children aged 0 to 6 years, irrespective of sex and ethnic origin. The intervention was any unlicensed or licensed MMRV regardless of administration route, dosage, and schedule. Outcomes related to immunogenicity and safety of vaccines had been reported.

Study Selection
Citations were merged together in Endnote, version X6, to facilitate management. Two reviewers independently applied the inclusion criteria to all retrieved articles and records of clinical trials in an unblinded standardized manner, evaluated by title, abstract, and full text. Disagreements between reviewers were resolved by consensus.

Data Collection
Data were mainly obtained from published articles. Unpublished clinical trials served as a supplement. Priority was given to data published in case of subtle discrepancy. For each of the eligible study, information of general study, study population, treatment, and outcomes of immunogenicity and/or safety was selectivity extracted onto piloted structured forms independently by 2 reviewers.

Quality Assessment
The internal validity of meta-analysis included clinical trials assessed using the Jadad score, 11 applying a score from 0 (very poor quality) to 5 (rigorous), based on the following items: randomization and adequate performance of the randomization procedure; double-blind and the adequate performance; and description of withdrawals and dropouts. Studies were graded on an ordinal scoring scale with higher scores representing studies of higher quality.

Statistical Analysis
Analysis of immunogenicity was performed mainly on according-to-protocol cohorts. Seroconversion or seroprotection rate and geometric mean titers (GMTs) for antibodies against each vaccine component after each dose were calculated with 95% confidence intervals (CIs) in each study. With respect to GMT, a log10 transformation was performed for the GMT to ensure normality, and the standard deviation (SD) was calculated from 95% CI using the calculator in RevMan 5.3 software (Cochrane Collaboration) to get completed continuous data. Analysis of safety was performed on per total vaccinated cohort. The incidences of solicited and/or unsolicited local, general symptoms, and adverse events were calculated.
Relative risks (RRs) for binary results and weighted mean difference (WMD) for continuous findings were calculated. Between-study heterogeneity was assessed by using Cochrane Q statistic and quantified by estimated I 2 . A Mantel-Haenszel fixed-effects model (M-H, fixed) for binary data and an inverse variance fixed-effects model (IV, fixed) for continuous data were used to calculate when the test for heterogeneity was not statistically significant (P > 0.10); otherwise, DerSimonian-Laird random-effects models (DL, random) were employed. [12][13][14] All statistical tests were 2-sided and considered significant when the P value was 0.05. We also performed sensitivity analyses to evaluate whether any single study dominated the results of meta-analyses. Finally, publication bias was assessed by quantitative Begg funnel plots for outcomes reported by the above 10 trials. 15 Statistical analyses were conducted using Stata 12.0 (StataCorp, College Station, Texas, USA) and RevMan 5.3.

Description of Studies Included
A total of 2190 studies and 99 clinical trial records were originally identified through online searching. The selection process is shown in Figure 1. Twenty-four articles published/ unpublished RCTs were included in the review and summarized in Table 1.  Considering the different controlled groups and vaccination intervals, only immunogenicity and safety profiles after first MMRV dose were compared and meta-analyzed. Of the 24 selected RCTs, 5 RCTs 24,27,28,37,39 were not included in the meta-analysis, because MMRV vaccine in these studies was given as second dose following a first dose of MMR or  Single MMRV Dose in Healthy Children Aged 9 to 24 Months

Immunogenicity
All included studies reported seroconversion rate as serological response outcome, except for varicella in Merck-MMRV-vaccinated studies, which was seroprotection rate. Seroconversion rate was defined as percent of subjects initially seronegative (with titers assay cut-offs), who developed postvaccination antibody titers above the assay cut-off levels. Seroprotection rate for varicella was defined as the proportion of subjects who were seronegative at baseline and whose postvaccination titer was !5 units/mL detected by the glycoprotein antigen-based enzyme-linked immunosorbent assay. The serological response rates and GMTs for all 4 antigens 27 to 84 days after vaccination were meta-analyzed in studies with comparison groups of MMR þ V, whereas 3 antigens (measles, mumps and rubella) for comparison groups of MMR. Considering the different laboratory assays and corresponding cut-offs (Table 1), only GSK-MMRV-vaccinated studies with same assays and cut-offs were combined for the analysis of GMT in this article.

Solicited Local Symptoms
Eight RCTs (MMRV vs MMR þ V) 19 Incidences of pain and swelling were around and below 10% in all groups, respectively. Grade 3 local reactions were rare (<0.5%) for all the vaccine groups, especially pain.

Solicited General Symptoms
Solicited general symptoms within 43 days (days 0-42) after vaccination from all the pooled analyses are presented in supplementary Table 1 (http://links.lww.com/MD/A491).
Fever was the most frequently reported solicited general symptom, pooled incidences of fever were around 60% in MMRV groups and 50% in MMR þ V/MMR groups. Majority were reported during the first 15 days (days 0-14) follow-up period. Half of the events were considered by the investigator to be related to investigational vaccine. Pooled incidence of grade 3 fever (rectal temperature >39.58C) during the 43 days after vaccination in these studies was relatively low (around 15% in MMRV groups, 11% in MMR þ V group and MMR group). Irrespective of follow-up period, higher incidences of fever were reported in MMRV group than in MMR þ V/MMR group (pooled RRs ranged from 1.12 to 1.60). All the analyses showed significant difference between comparison groups, except the analyses of related fever during 15 and 43 days after vaccination in MMRV versus MMR þ V comparison.

Serious Adverse Events (SAEs)
Incidences of any serious adverse events (SAEs) were around 1% in all the groups; only about one-tenth of the events were considered to be related to vaccination studied. About half of the related SAEs were febrile seizures. The incidence of related febrile seizure was under 0.8% in MMRV groups and under 0.5% in MMR þ V/MMR groups. No statistical difference was found between groups with no evidence of heterogeneity. No related fatal SAE was reported in any studies included.

MMRV administered as second dose after a first MMR/MMR R V/MMRV in healthy children
Children in 9 RCTs [20][21][22]25,26,29,31,33,34 received 2 doses of MMRV vaccines (Table 1). Actually, after administering 2 doses of MMRV in healthy children aged 9 to 24 months with an interval of 4 weeks to 6 months, MMRV showed strong immunogenicity against the 4 diseases. Moreover, MMRV was well tolerated compared with a dose of MMR followed by another dose of MMR, 29,31 MMR þ V, or MMRV, 33 and a dose of MMR þ V followed by another dose of MMR 22,26,34 or MMR þ V. 25 Second dose of MMRV ensured higher seroconversion rates (95%-100%) and GMTs for all vaccine components compared with the single vaccine dose schedule, especially up to 41.6-fold higher for antivaricella GMT. 33 Comparison between 2 doses of MMRV and MMR þ V showed antivaricella GMT increased 10.4-fold (95% CI 8.65, 12.41) in MMRV group and 5.2-fold (95% CI 4.14, 6.53) in MMR þ V group, compared with up to 2.2-fold increases for the GMTs of other 3 virus after the second dose in both groups. 25 Although 1 or 2 solicited local symptoms (pain, redness, and swelling) were more frequently reported after the second dose of MMRV compared to the first dose in most studies, 22,25,26,29,31,33,34 incidences of most adverse experiences for each comparison after the second dose were similar among groups.
Additionally, five RCTs suggested that MMRV vaccine seemed to be more immunogenic and well tolerated when given as a second dose after MMR (3 RCTs) 27,37,39 or MMR þ V (2 RCTs) 24,28 vaccination in children aged 15 months to 6 years ( Table 1).

Publication Bias
Publication bias was assessed by quantitative Begg funnel plots for 6 outcomes as the numbers of included studies were more than 10. No significant asymmetry was found.

DISCUSSION
This systematic review synthesized evidence about immunogenicity and safety of MMRV in 24 studies involving more than 23,000 healthy children. We used systematic strategy and broad search terms in multiple databases and related websites to identify as many published and unpublished clinical trials as possible. GMTs against the 4 diseases were assessed by performing a log10 transformation, to get a more intuitional understanding of the immunogenicity of vaccines. However, considering the different laboratory assays and corresponding cut-offs, only the GSK-MMRV-vaccinated studies with same assays and cut-offs were combined for the analyses of GMTs in this article. Moreover, considering the complex control designs and different intervals between doses, only the immunogenicity and safety of single MMRV dose were meta-analyzed in healthy children aged 9 to 24 months.
For immunogenicity, our results from the analysis of 19 RCTs suggested that single MMRV dose in healthy children aged 9 to 24 months had comparable immunogenicity profiles against these 4 diseases to MMR þ V/MMR. There are some exceptions as follows: (1) Antimeasles GMTwas significantly higher in MMRV group than that in MMR þV/MMR group, the GMT ratios were To improve individual protection against the 4 diseases and to have a more rapid impact on outbreaks, a second dose catch-up MMRV vaccination has been recommended. 1,6,40 Actually, after administering 2 doses of MMRV in healthy children aged 9 to 24 months with an interval ranging from 4 weeks to 6 months, MMRV showed strong immunogenicity against the 4 diseases. Second dose of MMRV ensured higher seroconversion rates (95%-100%) and GMTs for all vaccine components compared with the single vaccine dose schedule, especially for varicella virus. The efficacy of 2 doses of MMRV was also in line with estimates of 2 doses of MMR þ V, with higher immune response to the 4 virus components.
For safety, single MMRV dose in healthy children aged 9 to 24 months was generally well tolerated. Significant differences were demonstrated mainly in the comparisons of fever and rash. Fever was the most frequently reported solicited general symptom during the 43 days (days 0-42) of followup period in all included studies (pooled incidences were above 52.9% in all groups). Higher incidences of fever were found in MMRV groups compared to MMR þ V and MMR groups (RRs ranged from 1.12 to 1.60). Rash was the second frequently reported solicited general symptom. Generalized rash (RR ¼ 1.23; 95% CI 1.07, 1.40; P ¼ 0.004) and measles/rubella like rash (RR ¼ 1.44; 95% CI 1.15, 1.81; P ¼ 0.002) were significantly more frequent in MMRV group than in MMR þ V group. Moreover, measles/rubella-like rash (RR ¼ 1.45; 95% CI 1.06, 1.98; P ¼ 0.020) and varicella-like rash (RR ¼ 1.95; 95% CI 1.04, 3.66; P ¼ 0.040) were significantly more frequent in MMRV group than in MMR group. The results were consistent with a statistical modeling in a previous review, which indicated that the higher level of measles antibody titer after receipt of MMRV was positively associated with the higher rates of fever and measles-like rashes. 8 It is known that fever can precipitate febrile seizure in susceptible children aged 6 months to 5 years, especially in the 12 to 23-month age range. 41,42 The higher fever rate made the incidence of febrile seizure be more concerned. In several postmarketing observational safety surveillance studies, 43-46 an approximate 2-fold increase in risk for seizure or febrile seizure during 7 to 10 days or 5 to 12 days after vaccination were found among children aged 10 to 24 months, those who received the first dose of MMRV compared with those who received the first dose of MMR administered with or without varicella vaccine. However, no significant difference was found in our results in incidence of vaccine-related febrile seizure during the 43 days postvaccination of MMRV compared with MMR þ V/MMR. This might partly be due to the limited population, protocolspecified vaccinated ages, and intervals studied in RCTs. SAEs like febrile seizures that occur too infrequently to detect in RCTs may be identified and further studied through postmarketing restudies based on rigorous prospective study design.
Assessment of cost effectiveness based on a dynamic transmission model showed MMRV would provide more quality-adjusted life-years than MMR, and was cost-saving. 47 Economic analysis also showed that universal mass vaccination against varicella using MMRV to reduce the disease burden of varicella in Germany would lead to cost-savings from societal as well as from the health system perspective. 40 Considering the higher risk for seizure or febrile seizure, a model performed by Bauchau et al 48 suggested that use of MMRV instead of MMR þ V might substantially reduce number of hospitalizations despite the observed increased risk of febrile seizure, when MMRV was used as a first dose of measles-containing vaccine, which was one of the trade-offs between the two vaccination schemes. Our review highlights that providers who are considering administering MMRV to children should be concerned with the benefits and risks with parents or caregivers. Decisions should be made by them on a case-bycase basis.

CONCLUSIONS
In conclusion, this systematic review and meta-analysis showed rigorous evidence that MMRV had comparable immunogenicity and overall safety profiles to MMR administered with or without varicella vaccine.