Abstract
Background
Australian First Nations children are at very high risk of early, recurrent, and persistent bacterial otitis media and respiratory tract infection. With the PREVIX randomised controlled trials, we aimed to evaluate the immunogenicity of novel pneumococcal conjugate vaccine (PCV) schedules.
Methods
PREVIX_BOOST was a parallel, open-label, outcome-assessor-blinded, randomised controlled trial. Aboriginal children living in remote communities of the Northern Territory of Australia were eligible if they had previously completed the three-arm PREVIX_COMBO randomised controlled trial of the following vaccine schedules: three doses of a 13-valent PCV (PCV13; PPP) or a ten-valent pneumococcal Haemophilus influenzae protein D conjugate vaccine (PHiD-CV10; SSS) given at 2, 4, and 6 months, or SSS given at 1, 2, and 4 months followed by PCV13 at 6 months (SSSP). At age 12 months, eligible children were randomly assigned by a computer-generated random sequence (1:1, stratified by primary group allocation) to receive either a PCV13 booster or a PHiD-CV10 booster. Analyses used intention-to-treat principles. Co-primary outcomes were immunogenicity against protein D and serotypes 3, 6A, and 19A. Immunogenicity measures were geometric mean concentrations (GMC) and proportion of children with IgG concentrations of 0·35 μg/mL or higher (threshold for invasive pneumococcal disease), and GMCs and proportion of children with antibody levels of 100 EU/mL or higher against protein D. Standardised assessments of otitis media, hearing impairment, nasopharyngeal carriage, and developmental outcomes are reported. These trials are registered with ClinicalTrials.gov (NCT01735084 and NCT01174849).
Findings
Between April 10, 2013, and Sept 4, 2018, 261 children were randomly allocated to receive a PCV13 booster (n=131) or PHiD-CV10 booster (n=130). Adequate serum samples for pneumococcal serology were obtained from 127 (95%) children in the PCV13 booster group and 126 (97%) in the PHiD-CV10 booster group; for protein D, adequate samples were obtained from 126 (96%) children in the PCV13 booster group and 123 (95%) in the PHiD-CV10 booster group. The proportions of children with IgG concentrations above standard thresholds in PCV13 booster versus PHiD-CV10 booster groups were the following: 71 (56%) of 126 versus 81 (66%) of 123 against protein D (difference 10%, 95% CI −2 to 22), 85 (67%) of 127 versus 59 (47%) of 126 against serotype 3 (−20%, −32 to −8), 119 (94%) of 127 versus 91 (72%) of 126 against serotype 6A (−22%, −31 to −13), and 116 (91%) of 127 versus 108 (86%) of 126 against serotype 19A (−5%, −13 to 3). Infant PCV13 priming mitigated differences between PCV13 and PHiD-CV10 boosters. In both groups, we observed a high prevalence of otitis media (about 90%), hearing impairment (about 75%), nasopharyngeal carriage of pneumococcus (about 66%), and non-typeable H influenzae (about 57%). Of 66 serious adverse events, none were vaccine related.
Interpretation
Low antibody concentrations 6 months post-booster might indicate increased risk of pneumococcal infection. The preferred booster was PCV13 if priming did not have PCV13, otherwise either PCV13 or PHiD-CV10 boosters provided similar immunogenicity. Mixed schedules offer flexibility to regional priorities. Non-PCV13 serotypes and non-typeable H influenzae continue to cause substantial disease and disability in Australian First Nation's children.
