Sustained and indirect effects of PCV10 reduced-dose schedules on pneumococcal carriage in Viet Nam: a long-term follow-up of a cluster-randomised controlled trial.

Abstract

A cluster-randomised trial in Nha Trang, Viet Nam, previously showed non-inferiority of a reduced two-dose (1p + 1) pneumococcal conjugate vaccine 10 (PCV10) schedule compared with three-dose schedules (2p + 1, 3p + 0) for controlling vaccine-type (VT) nasopharyngeal carriage after 3·5 years, but the long-term sustainability of reduced-dose schedules and their indirect effects on unvaccinated age groups remain incompletely understood. We aimed to assess the sustained and indirect effects of reduced-dose PCV10 schedules on VT carriage 5·5 years after vaccine introduction. This study presents long-term follow-up data from a cluster-randomised controlled trial conducted in 24 communes in Nha Trang, Viet Nam. These communes were randomly assigned (1:1:1:1; six communes per group) by automated rejection sampling to one of four PCV10 vaccination schedules for infants: 0p + 1 (single dose at age 12 months), 1p + 1 (doses at ages 2 and 12 months), 2p + 1 (doses at ages 2, 4, and 12 months), or 3p + 0 (doses at ages 2, 3, and 4 months); three unvaccinated communes provided observational comparison. The prespecified primary outcome for this follow-up study was non-inferiority of 1p + 1 versus standard-dose (2p + 1, 3p + 0) schedules for prevalence of VT carriage in vaccine-eligible infants (aged 4-11 months) and toddlers (aged 14-24 months) in 2022 (5·5 years after vaccine introduction. A prespecified secondary objective evaluated the 0p + 1 schedule. Indirect effects on VT carriage prevalence in adult caregivers and preschool children (aged 3-4 years) were also assessed. Non-inferiority was defined as the upper bound of the 95% CI of the difference in VT carriage prevalence not exceeding 5 percentage points (pp). Age-specific trends in VT carriage proportion over the trial period were estimated using Poisson regression models with intervention groups pooled. This trial was completed and is registered with ClinicalTrials.gov, NCT02961231. Between Oct 10, 2016, and Aug 23, 2022, 49 644 participants were enrolled, including 10 423 infants (aged 4-11 months), 10 988 toddlers (aged 14-24 months), 9580 preschool children (aged 3-4 years), and 18 653 caregivers. Among the 30 991 children, 15 018 (48·5%) were female and 15 973 (51·5%) were male. 27 384 (88·4%) were enrolled in vaccinated communes (0p + 1: 6984, 1p + 1: 6906, 2p + 1: 6972, 3p + 0: 6522) and 3607 (11·6%) in the unvaccinated communes. At 5·5 years, prevalence of PCV10-type carriage in infants was 0·7% (2/286) in the 1p + 1 group, 1·9% (6/314) in the 2p + 1 group, and 0·9% (2/226) in the 3p + 0 group; in toddlers, it was 0·9% (3/332) in the 1p + 1 group, 0·6% (2/344) in the 2p + 1 group, and 1·0% (3/300) in the 3p + 0 group. The 1p + 1 schedule remained non-inferior to three-dose schedules in all infant and toddler comparisons (differences in infants of -1·2 pp [95% CI -3·0 to 0·6] for 1p + 1 vs 2p + 1 and -0·2 pp [-1·7 to 1·4] for 1p + 1 vs 3p + 0; differences in toddlers of 0·3 pp [-1·0 to 1·6] for 1p + 1 vs 2p + 1 and -0·1 pp [-1·6 to 1·4] for 1p + 1 vs 3p + 0). The 0p + 1 schedule met non-inferiority in seven of eight intention-to-treat comparisons. The proportion of pneumococcal carriers carrying PCV10-type serotypes declined in infants (from 52·1% [160/307] to 7·6% [12/157]), toddlers (from 50·0% [246/492] to 4·0% [15/378]), and adult caregivers (from 39·4% [28/71]) to 10·5% [12/114]), at similar rates across age groups. During the trial period, 50 serious adverse events were reported within 1 month after vaccination, none of these events was deemed related to PCV10 vaccination. All children recovered without sequelae. The 1p + 1 schedule remained non-inferior to standard three-dose schedules after 5·5 years. Across all intervention groups, VT carriage proportion declined at comparable rates in vaccinated children and unvaccinated adult caregivers, demonstrating that reduced-dose schedules can generate population-level indirect protection similar to standard schedules. These findings are most applicable to settings with established pneumococcal conjugate vaccine programmes following a catch-up campaign, and support 1p + 1 as a viable strategy for maintaining pneumococcal disease control while reducing programmatic costs and delivery complexity. Gates Foundation. For the Vietnamese translations of the abstract see Supplementary Material section.