Global impact of ten-valent and 13-valent pneumococcal conjugate vaccines on invasive pneumococcal disease in all ages (the PSERENADE project): a global surveillance analysis

Abstract

Background: Pneumococcal conjugate vaccines (PCVs) that are ten-valent (PCV10) and 13-valent (PCV13) became available in 2010. We evaluated their global impact on invasive pneumococcal disease (IPD) incidence in all ages.

Methods: Serotype-specific IPD cases and population denominators were obtained directly from surveillance sites using PCV10 or PCV13 in their national immunisation programmes and with a primary series uptake of at least 50%. Annual incidence rate ratios (IRRs) were estimated comparing the incidence before any PCV with each year post-PCV10 or post-PCV13 introduction using Bayesian multi-level, mixed-effects Poisson regressions, by site and age group. All site-weighted average IRRs were estimated using linear mixed-effects regression, stratified by product and previous seven-valent PCV (PCV7) effect (none, moderate, or substantial).

Findings: Analyses included 32 PCV13 sites (488 758 cases) and 15 PCV10 sites (46 386 cases) in 30 countries, primarily high income (39 sites), using booster dose schedules (41 sites). By 6 years after PCV10 or PCV13 introduction, IPD due to PCV10-type serotypes and PCV10-related serotype 6A declined substantially for both products (age <5 years: 83-99% decline; ≥65 years: 54-96% decline). PCV7-related serotype 19A increases before PCV10 or PCV13 introduction were reversed at PCV13 sites (age <5 years: 61-79% decline relative to before any PCV; age ≥65 years: 7-26% decline) but increased at PCV10 sites (age <5 years: 1·6-2·3-fold; age ≥65 years: 3·6-4·9-fold). Serotype 3 IRRs had no consistent trends for either product or age group. Non-PCV13-type IPD increased similarly for both products (age <5 years: 2·3-3·3-fold; age ≥65 years: 1·7-2·3-fold). Despite different serotype 19A trends, all-serotype IPD declined similarly between products among children younger than 5 years (58-74%); among adults aged 65 years or older, declines were greater at PCV13 (25-29%) than PCV10 (4-14%) sites, but other differences between sites precluded attribution to product.

Interpretation: Long-term use of PCV10 or PCV13 reduced IPD substantially in young children and more moderately in older ages. Non-vaccine-type serotypes increased approximately two-fold to three-fold by 6 years after introduction of PCV10 or PCV13. Continuing serotype 19A increases at PCV10 sites and declines at PCV13 sites suggest that PCV13 use would further reduce IPD at PCV10 sites.

Funding: Bill & Melinda Gates Foundation as part of the WHO Pneumococcal Vaccines Technical Coordination Project.

Figure 1

Sites that contributed IPD incidence data to the PSERENADE project and reasons for exclusion from present analysis IPD=invasive pneumococcal disease. PCV=pneumococcal conjugate vaccine. PCV10=ten-valent PCV (Synflorix, GSK). PCV13=13-valent PCV (Prevenar 13, Pfizer).

Figure 2

All-site weighted average IRRs for PCV10-type IPD by age group Site-specific annual incidence rates of IPD following PCV10 or PCV13 introduction were compared to the site's average incidence rate before PCV introduction to compute site-specific IRRs; all-site-weighted average IRRs are an average of the sites' IRRs. PCV10 types include serotypes 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, and 23F. Cross-protection from serotype 6A was not included. The shaded area indicates the year of PCV10 or PCV13 rollout. Estimates at year 0 indicate the change in incidence after the first year of PCV10 or PCV13 use. In individuals aged 5 years or older, the model that best fit the data was one that shifted the first year of estimated effect to the year after the first year of PCV introduction, not the year of PCV rollout as for children younger than 5 years. For sites with moderate or substantial PCV7 impact, the first year of estimated effect occurred at PCV7 introduction (not shown). For sites with no PCV7 impact, the first year of estimated effect can be seen for the first point estimate, indicating the pre-PCV period (ie, IRR=1), which occurs at year –1 for children younger than 5 years and at year 0 for all other age groups. IPD=invasive pneumococcal disease. IRR=incidence rate ratio. PCV=pneumococcal conjugate vaccine. PCV7=seven-valent PCV (Prevenar, Pfizer). PCV10=ten-valent PCV (Synflorix, GSK). PCV13=13-valent PCV (Prevenar 13, Pfizer).

Figure 3

All-site weighted average IRRs for IPD due to any serotype or specific serotypes or serotype groups for children younger than 5 years Site-specific annual incidence rates of IPD following PCV10 or PCV13 introduction were compared to the site's average incidence rate before PCV introduction to compute site-specific IRRs; all-site weighted average IRRs are an average of the sites' IRRs. The y-axis scale changes between figures. PCV13 types include serotypes 1, 3, 4, 5, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F, and 23F. The shaded area indicates the year of PCV10 or PCV13 rollout. Estimates at year 0 indicate the change in incidence after the first year of PCV10 or PCV13 use. In individuals aged 65 years or older, the model that best fit the data was one that shifted the first year of estimated effect to the year after the first year of PCV introduction, not the year of PCV rollout as for children younger than 5 years. For sites with moderate or substantial PCV7 impact, the first year of estimated effect occurred at PCV7 introduction (not shown). For sites with no PCV7 impact, the first year of estimated effect can be seen for the first point estimate, indicating the pre-PCV period (ie, IRR=1), which occurs at year –1 for children younger than 5 years and at year 0 for all other age groups. Similar figures for age groups 5–17 years, 18–49 years, and 50–64 years are in appendix 2 (pp 23–25). IPD=invasive pneumococcal disease. IRR=incidence rate ratio. PCV=pneumococcal conjugate vaccine. PCV7=seven-valent PCV (Prevenar, Pfizer). PCV10=ten-valent PCV (Synflorix, GSK). PCV13=13-valent PCV (Prevenar 13, Pfizer).

Figure 4

All-site weighted average IRRs for IPD due to any serotype or specific serotypes or serotype groups for adults aged 65 years or older Site-specific annual incidence rates of IPD following PCV10 or PCV13 introduction were compared to the site's average incidence rate before PCV introduction to compute site-specific IRRs; all-site weighted average IRRs are an average of the sites' IRRs. The y-axis scale changes between figures. PCV13 types include serotypes 1, 3, 4, 5, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F, and 23F. The shaded area indicates the year of PCV10 or PCV13 rollout. Estimates at year 0 indicate the change in incidence after the first year of PCV10 or PCV13 use. In individuals aged 65 years or older, the model that best fit the data was one that shifted the first year of estimated effect to the year after the first year of PCV introduction, not the year of PCV rollout as for children younger than 5 years. For sites with moderate or substantial PCV7 impact, the first year of estimated effect occurred at PCV7 introduction (not shown). For sites with no PCV7 impact, the first year of estimated effect can be seen for the first point estimate, indicating the pre-PCV period (ie, IRR=1), which occurs at year –1 for children younger than 5 years and at year 0 for all other age groups. Similar figures for age groups 5–17 years, 18–49 years, and 50–64 years are in appendix 2 (pp 23–25). IPD=invasive pneumococcal disease. IRR=incidence rate ratio. PCV=pneumococcal conjugate vaccine. PCV7=seven-valent PCV (Prevenar, Pfizer). PCV10=ten-valent PCV (Synflorix, GSK). PCV13=13-valent PCV (Prevenar 13, Pfizer).