Abstract

Background: Pneumococcal conjugate vaccines (PCVs) provide direct protection against disease in those vaccinated, and interrupt transmission through the prevention of nasopharyngeal carriage. Methods: We analysed immunogenicity data from 5224 infants who received PCV in prime-boost schedules. We defined any increase in antibody between the one-month post-priming visit and the booster dose as an indication of nasopharyngeal carriage ('seroincidence'). We calculated antibody concentrations using receiver-operator characteristic curves, and used generalised additive models to compute their protective efficacy against seroincidence. To support seroincidence as a marker of carriage, we compared seroincidence in a randomised immunogenicity trial in Nepal with the serotype-specific prevalence of carriage in the same community. Findings: In Nepalese infants, seroincidence of carriage closely correlated with serotype-specific carriage prevalence in the community. In the larger data set, antibody concentrations associated with seroincidence were lowest for serotypes 6B and 23F (0.50 µg/mL and 0.63 µg/mL respectively), and highest for serotypes 19F and 14 (2.54 µg/mL and 2.48 µg/mL respectively). The protective efficacy of antibody at these levels was 62% and 74% for serotypes 6B and 23F, and 87% and 84% for serotypes 19F and 14. Protective correlates were on average 2.15 times higher in low/lower middle income countries than in high/upper middle income countries (GMR 2.15, 95%CI 1.46-3.17, p=0.0024). Interpretation: Antibody concentrations associated with protection vary between serotypes. Higher antibody concentrations are required for protection in low-income countries. These findings are important for global vaccination policy, to interrupt transmission by protecting against carriage.