Immunogenicity data available from Australian children included in a Phase II randomized controlled study4
(NCT00134719) and US children enrolled in a Phase III randomized controlled study5
(NCT00289783) were assessed. The studies, which were described in detail previously,4,5
were designed in a similar way within the context of each country’s immunization schedule, with the pre-specified objective of pooling data for the primary objective of assessing the non-inferiority of MMR and VAR co-administered with Hib-MenCY compared with co-administration with a licensed Hib vaccine. Conditions to pool the data were therefore defined prospectively, and were determined on the grounds of comparable immunogenicity between studies and were met if the point estimates of the difference between the HibMenCY-TT and Hib-OMP groups were above the predefined noninferiority limits in each study in terms of anti-measles seroconversion, anti-mumps seroconversion, anti-rubella seroresponse, and anti-varicella seroconversion.
Immunogenicity analyses were performed on the pooled according-to-protocol (ATP) cohorts for immunogenicity from each study, defined as vaccinated children who met all eligibility criteria, complied with protocol-defined procedures, and had antibody assay results available for at least 1 antigen component. Cut-offs for seropositivity were defined as follows: anti-measles, enzyme-linked immunosorbent assay (ELISA) value ≥ 150 mIU/mL; anti-mumps, neutralization test value ≥ 28 ED50; anti-rubella, ELISA value ≥ 4 IU/mL; anti-varicella fluorescent antibody membrane assay value ≥ 1:5. Seroconversion for antibodies to measles, mumps, and varicella was defined as the presence of detectable levels of the relevant antibodies at 42 d post-vaccination in subjects who were seronegative before vaccination. Seroresponse for rubella was defined as the appearance of anti-rubella virus antibodies to a concentration ≥ 10 IU/mL in subjects who were seronegative before vaccination (concentration < 4 IU/mL).
The coprimary objectives of the pooled immunogenicity analysis were to demonstrate noninferiority of MMR and VAR when coadministered with a dose of HibMenCY-TT compared with MMR and VAR coadministered with a dose of Hib-OMP in terms of anti-measles seroconversion, anti-mumps seroconversion, anti-rubella seroresponse, and anti-varicella seroconversion at 42 d (range 35–56 d) after administration of the vaccines. The pooling criterion prespecified in the study protocols stated that the data could be pooled if, in each individual study, the point estimate for the group difference in seroconversion/seroresponse rates (HibMenCY-TT group minus Hib-OMP group) was ≥ -5% for measles, mumps, and rubella, and ≥ -10% for varicella (note: the individual studies were not adequately powered to demonstrate a lower limit higher than the noninferiority criterion). Assessment of noninferiority was based on standardized asymptotic 95% confidence intervals (CIs) for the difference between groups (HibMenCY-TT group minus Hib-OMP control group) in percentage of subjects reaching the separate prespecified endpoints for each of the MMR and VAR components. Noninferiority would be demonstrated if the lower limit of the 95% CI was above -5% for measles, mumps, and rubella, and above -10% for varicella.
Parents/guardians of subjects completed diary cards with any solicited symptoms specific to MMR and VAR vaccination: fever ≥ 38°C measured by any route, rash/exanthem, parotid/salivary gland swelling, and any suspected meningeal signs or febrile convulsions over a 43-d follow-up period after vaccination. Subjects with symptoms of parotid swelling were to return to the study center to be medically examined to confirm the diagnosis of parotiditis. As planned in the study protocols, the reactogenicity and safety analysis of these symptoms was performed on subjects in the pooled total vaccinated cohorts (TVCs) from both studies with safety data available (i.e., diary cards returned). Relative risk (percentage of subjects reporting a specific symptom in HibMenCY-TT group over percentage in Hib-OMP group) was calculated with exact 95% CI for each symptom. P-values were calculated using a 2-sided Exact Stratified Test conditional to number of cases and taking into account the study effect.