Reply to Dr. Peter Paradiso

Dear Editor,

Dr. Paradiso states that, although 60-70% of US adults have been vaccinated, “there has been no measurable impact on invasive pneumococcal disease, let alone pneumonia, caused by the serotypes in the vaccine.” This is not correct. If he means a population-wide effect, pneumococcal polysaccharide vaccine (PPSV23) has not been studied in this fashion in the US. The CDC ABC’s didn’t begin to report data until PPSV23 was in widespread use and only a few years before the introduction of conjugate vaccine (PCV7) into the pediatric population. However, Andrews et al.² documented a population-wide impact of vaccination with PPSV23 in England and Wales, although it was, admittedly, not as great as might have been expected from earlier studies.

The absence of data is not the same as the absence of effect. A steady and substantial decline in the proportion of cases of community acquired pneumonia (CAP) attributable to Streptococcus pneumoniae in the decades leading up to 2000 has paralleled the increased use of PPSV23. Furthermore, the US, which had a higher rate of PPSV23 use, has a much lower rate of pneumococcal disease than do European countries; recent studies in the US show that pneumococcus causes <10% of CAP.^3,4 These observations are, of course, not necessarily causally related: decreased rates of cigarette smoking, increased uptake of influenza vaccination and many other factors, recognized or unrecognized, may, of course, play a role. The increase in non-PCV7 serotypes after implementation of universal PCV7 vaccination might have been much greater had not 60-70% of adults received PPSV23.

Regarding duration of a vaccine effect, it is difficult to compare the UK and Dutch studies. Shapiro et al.⁵ provided beautiful epidemiological data on this same point, showing >50% protection from polysaccharide pneumococcal vaccine effect beyond 5 y except in the most elderly subjects studied. It is further worth noting that Jackson et al.⁶ showed a complete loss of opsonophagocytic effect one year after vaccination with either PCV13 or PPSV23.

Most importantly, we see 2 very large and different elephants in the room. (1) National use of PCV7 nearly eliminated disease in adults caused by the 7 vaccine serotypes, and we have no reason to believe that the same effect will not be observed in the case of the 13 serotypes in PCV13. Thus we continue to regard a campaign to vaccinate all adults >65 years of age with PCV13 as well-intended but misguided. (2) The CAPiTA study very clearly failed to show any benefit from PCV13 in persons who became immune compromised after they had enrolled (immunocompromise was an exclusion criterion for initial enrollment). As a result, we similarly question the ACIP recommendation for vaccination of all immunocompromised adults ages 19–64.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank our friend and colleague, Dr. Paradiso, for his interest in our recent article.¹