Preparing for Next-Generation Vaccines: Breakthrough Invasive Pneumococcal Disease Patterns Post-PCV13 and Post-PPSV23 in Australia

Pneumococcal conjugate (PCV) and polysaccharide (PPSV) vaccines have reduced invasive pneumococcal disease (IPD) in Australia. We evaluated breakthrough IPD following PCV13 and PPSV23 immunization (2009-2023) to inform next-generation vaccine strategies.

We analyzed laboratory-confirmed IPD cases from the National Notifiable Diseases Surveillance System to estimate the vaccine-type proportion among vaccinated IPD cases (those meeting the minimum age-specific dose threshold), whose vaccine-type infections constitute breakthrough IPD. Temporal trends were assessed using logistic regression, and subgroup differences using chi-square or Fisher's tests. Random Forest and Gradient Boosting models with five-fold cross-validation classified IPD due to PPSV23-type serotypes.

Across 25,643 IPD cases, annual notifications increased while PCV13-type proportions declined. PCV13-type proportions were 19.4% in infants and 37.5% at 2-14 years; PPSV23-type proportions were 47.5% at 15-69 and 48.4% at ≥70 years, with serotype 3 the largest contributor. Vaccinated individuals showed lower vaccine-type proportions than non- or partially vaccinated cases across age groups. Vaccine-type proportions were highest for pneumonia and in the Northern Territory, with greatest residual burden among older adults and Indigenous Australians. Machine-learning models classifying PPSV23-type IPD (accuracy ≈92%) identified vaccination-related variables, clinical presentation, and year of infection as key predictors; excluding these variables reduced AUC from 0.97-0.98 to 0.69.

PCV13-type proportions declined over time, most markedly in infants, whereas PPSV23-type proportions remained persistently high in older adults and Indigenous Australians despite higher documented vaccination. Serotype 3, included in all current conjugate formulations, dominated breakthrough for both vaccines, and in descriptive comparisons limited by small subgroup sizes, neither additional doses nor sequential vaccination was associated with lower vaccine-type proportions.