Long-term population impact of infant 10-valent pneumococcal conjugate vaccination on invasive pneumococcal disease in adults in Finland.

BACKGROUND

Limited data are available on long-term indirect effects of ten-valent pneumococcal conjugate vaccine (PCV10) programmes. We evaluated changes in invasive pneumococcal disease (IPD) incidence, mortality, and serotype distribution in adults up to 9 years after infant PCV10 introduction.

METHODS

Culture-confirmed IPD cases ≥18 years (n = 5610; 85% were pneumonia) were identified through national, population-based laboratory surveillance; data were linked with population registry to conduct nationwide follow-up study. In a time-series model, we compared serotype-specific IPD incidence and associated 30-day mortality rates before and after PCV10 by using negative binomial regression models.

RESULTS

During pre-PCV10 period (7/2004-6/2010), overall IPD incidence in adults ≥18 years increased yearly by 4.8%. After adjusting for trend and seasonality, the observed PCV10 serotype IPD incidence in 7/2018-6/2019 was 90% (12/100,000 person-years) lower than the expected rate without PCV10 program. Non-PCV10 serotype incidence was 40% (4.4/100,000 person-years) higher than expected; serotypes 3, 19A, 22F, and 6C accounted for most of the rate increase. However, incidence of non-PCV10 IPD levelled off by end of follow-up. The observed-expected incidence rate-ratio (IRR) was 0·7 (95 %CI 0·5-0.8) for all IPD and 0·7 (95 %CI 0·3-1·3) for IPD-associated 30-day mortality. Case-fatality proportion decreased from 11·9% to 10.0% (p < 0.01). In persons ≥65 years, the IRR was 0·7 (95 %CI 0·5-0.95).

CONCLUSIONS

Significant indirect effects were seen for vaccine-serotype IPD and for overall IPD in all adult age groups. For non-vaccine IPD, the incidence stabilized 5 years after infant PVC10 program introduction, resulting in a steady state in which non-vaccine IPD accounted for nearly 90% of overall IPD. Substantial pneumococcal disease burden remains in older adults.