13-valent Pneumococcal Conjugate Vaccine Efficacy is Declining with Old Age: Results from an Exploratory Analysis of the CAPiTA Trial
Background: Immunogenicity of vaccines may be impaired in adults of high age. We retrospectively analyzed the impact of age on the efficacy of 13-valent pneumococcal conjugate vaccine (PCV13) in the prevention of vaccine-type community-acquired pneumonia (VT-CAP) in healthy subjects aged 65 years and older using data from the CAPiTA trial.
Methods: The CAPiTA trial was a randomized controlled study where 84,496 subjects aged 65 and over were randomized to PCV13 or placebo. The effects of vaccination, age and a vaccine-age interaction term were assessed using Cox proportional hazard models, with the first episode of X-ray confirmed VT-CAP as the outcome. Relative vaccine efficacy (VE) by age was calculated as 1 – hazard ratio (HR) and absolute reduction of VT-CAP was calculated from the model-predicted incidences.
Results: The median age was 71 years (IQR 68-76), and 47,252 (56%) were male. During a mean follow-up period of 4 years, 139 first episodes of VT-CAP were recorded. The vaccine-age interaction was statistically significant (p=0.036), with the HR of vaccination increasing with 6% per year (Table 1). The model-predicted VE declined from 69% (95%CI 29-100%) in 65 year olds to 44% (17-71%) in 75 year olds (Figure 1). Absolute reduction of VT-CAP incidence was declining less rapidly (Figure 2). There were too few cases in subjects above 80 years of age to determine VE in this age group. In sensitivity analysis with subjects aged < 85 years, which had 130 episodes of VT-CAP, the interaction effect was attenuated to 3.6% per year and not statistically significant, suggesting that the size of the interaction effect may have been overestimated caused by the few VT-CAP episodes in the highest age group in PCV13 vaccinated subjects.
Conclusion: The model predicted a decrease in VE with increasing age. Since this trial was not designed to study vaccine-age interaction, and number of cases in the oldest age groups were too few to draw VE conclusions, observational studies should be considered to evaluate this question further. These findings may be relevant for policy makers, vaccination program managers, and health workers.
Table 1: Vaccine-age interaction
age - 65
vaccine * (age - 65)
Figure 1: Model derived vaccine efficacy
Figure 2: Absolute reduction of VT-CAP incidence
C. H. Van Werkhoven,
M. Bolkenbaas, None
C. Webber, Pfizer Vaccines Clinical Research: Employee and Shareholder, Salary
B. Schmoele-Thoma, Pfizer Vaccines Clinical Research: Employee and Shareholder, Salary
S. D. Patterson, Pfizer Vaccines Clinical Research: Employee, Salary
W. Gruber, Pfizer Vaccines Clinical Research: Employee and Shareholder, Salary
D. E. Grobbee, Pfizer Vaccines Clinical Research: Investigator, Research grant
M. Bonten, Pfizer Vaccines Clinical Research: Investigator, Research grant