Response to Mungall et al. letter to the editor on Streptococcus pneumoniae serotype 19A: worldwide epidemiology. Expert review of vaccines 2017;16(10):1007–27

We thank Mungall and colleagues for their interest in our manuscript that focused on serotype 19A. The letter addresses a diversity of issues beyond the original topic of our article. It also contains misinterpretation of the available literature. We address the most relevant:

1. ‘The cost-effectiveness results that Isturiz et al. allude to are also based on inaccurate assumptions around 19A efficacy for PHiD-CV as well as those around indirect protection and acute otitis media efficacy, and we refer the reader to a previous correction of these assumptions’ [1].

Reference [1] as cited by Mungall et al. was a letter to the editor in response to a cost-effectiveness publication by Wu and colleagues in 2016 [2]. This paper by Wu et al. was not referenced in our article. In addition, as data continuously emerge, cost-effectiveness of PCVs should be evaluated in the context of the most current data. We therefore refer the readers to a full and more recent review of the vaccine effectiveness (VE) assumptions used in economic evaluation of infant vaccination programs [3]. This review shows that the previous assumptions related to 19A effectiveness for PCV10 as well as those around indirect protection and ability to prevent acute otitis media are not consistent with the current body of evidence and may lead to spurious results in cost-effectiveness evaluations published in recent years.

2. ‘PCV13 VE against serotype 3 has not been consistently demonstrated, with serotype-specific VE estimates reported by the majority of studies being not significantly different from zero.’

Our article focused on serotype 19A, and made no reference to VE for serotype 3. However, in the interest of scientific accuracy, we think it essential to review the data. In summary, the above statement is inaccurate. The single primary reference cited by Mungall et al. is a study which was conducted in Catalonia, Spain [4]. Catalonia did not introduce PCV13 into the recommended program until January 2017. In this study, the reported statistical power for the VE estimate for serotype 3 invasive pneumococcal disease (IPD) in children for ≥1 dose was 23%. We conducted a systematic review of the literature and identified six published indirect cohort or case-control studies reporting the effectiveness of PCV13 against IPD associated with serotype 3 in children, using either at 2 + 1 or 3 + 1 regimen, including the study conducted in Spain [4–9]. When considering those studies with nonoverlapping datasets, all showed a positive VE estimate which ranged from 26% (2 + 1 regimen) to 80% (3 + 1 regimen). In addition, the Streptococcus pneumoniae Invasive Disease network (SpIDnet), which has been funded by the European Centre for Disease Prevention and Control (ECDC) since 2012 to perform active population-based surveillance of IPD in children across the European Union, has assessed PCV13 VE for serotype 3 [10,11]. In the largest indirect cohort analysis of PCV13 effectiveness reported to date, including data across nine sites in the European Union, the PCV13 VE (adjusted for site, age, sex, underlying conditions, and year of notification) for serotype 3 IPD was 70% (95% confidence interval [CI]: 44–83) for at least one dose of PCV13 and 57% (95% CI: 5–81) for children who were fully vaccinated. The ECDC also recently reported the indirect effect of childhood PCV vaccination programs in the elderly across 13 sites in the European Union [12]. In sites with universal PCV10 vaccination, the incidence of serotype 3 IPD in adults aged ≥65 years increased in all sites (pooled increase of 58% in 2015 compared to 2010). In sites with universal PCV13 vaccination, the incidence of serotype 3 IPD in adults showed a pooled decrease of 11%.

Scientific rigor requires that decisions about the utility of an intervention or the validity of a hypothesis not be based on the results of a single study, as results may vary from one study to the next and the dynamics of serotypes may vary per patient, season, calendar year, and geography. Rather, decisions on the utility of an intervention should be based on the totality of the evidence and such evidence demonstrates PCV13 provides direct protection for serotype 3.

3. ‘Surveillance data from countries using PHiD-CV indicate some variability in serotype 19A observations but does not suggest this is markedly different from what is seen with PCV13 use’.

The above statement also is inaccurate and refuted by the totality of available evidence worldwide. In the aforementioned study by the ECDC, surveillance data from sites with universal PCV13 vaccination found that the incidence of serotype 19A IPD in adults aged ≥65 years declined in all sites (pooled decline of 40% in 2015). By contrast, in sites with universal PCV10 programs, serotype 19A increased in all sites (227% pooled increase in 2015).

Finland, the Netherlands, Brazil, and Belgium all use PCV10 in their infant immunization programs. The most recent national population-based surveillance data from Finland, not included in the response by Mungall et al., found no net reduction in 19A IPD in the vaccine-eligible cohort following 6 years of PCV10 use [13]. Recent reports from the Netherlands and Brazil came to a similar conclusion [14,15]. Furthermore, the most recent national surveillance data from Belgium document that after near-elimination of 19A in Belgian children in 2016 following PCV13 use in the pediatric national immunization program (NIP), a 10-fold increase in 19A IPD cases in children below 2 years of age was observed in 2017 (Figure 1) following a switch to PCV10 in the NIP during June 2015 and April 2016; the Belgium National Reference Laboratory for Pneumococci has attributed this increase to ‘the re-emergence of serotype 19A’ [16]. This increase should be interpreted with caution, since it represents data from one surveillance season and could be due to mechanisms other than a switch in vaccine, including year-to-year shifts in pneumococcal serotype distribution, antibiotic use, or other factors.

Figure 1. Number of IPD cases in children <2 years of age in Belgium. Adapted from the National Reference Laboratory for Pneumococci. Surveillance van de pneumokokkeninfecties in België. Verslag voor 2017. [Accessed 05.05.2018]; Available from: https://nrchm.wiv-isp.be/nl/ref_centra_labo/streptococcus_pneumoniae_invasive/Rapporten/Streptococcus%20pneumoniae%202,017.pdf.

While the incidence of 19A IPD in children has remained low in some PCV10 countries, we know of no PCV10 country that has experienced an absolute reduction of 19A IPD in children or adults following PCV10 use and many countries have experienced an increase, in contrast to countries using PCV13 in their NIP. A lack of direct protection afforded by PCV10 for serotype 19A had been recognized by several public health authorities, as summarized in our review.

Finally, Mungall et al. suggest that the more relevant approach to analyzing the public health benefits of pneumococcal vaccination would be to evaluate the impact of vaccines on overall disease. While this argument is appealing, it is essentially a straw man because such a direct comparison is an impossible bar to achieve for many reasons: (1) pneumococcal disease incidence will be influenced by many factors other than vaccine, including natural fluctuations in serotypes, antibiotic use, different demographic characteristics such as population age, use of influenza or pneumococcal polysaccharide vaccines, breastfeeding and group childcare practices, cigarette smoking practices, household crowding, air pollution, and others; (2) the need to include the totality of outcomes including IPD, pneumonia, otitis media, sinusitis, antibiotic resistance, and others; (3) the multiyear benefit that many PCV10 countries have experienced from seven of the components in PCV13 through prior use of PCV7; and (4) the absence of and the near impossibility to conduct an adequately powered trial directly comparing the two vaccines. It also may be a scientifically invalid approach, as it is based on the idea that the serotype mix within overall IPD is irrelevant. However, serotypes may differ in predilection for different disease syndromes within IPD (e.g. meningitis versus bacteremia), invasiveness, age distribution, and antibiotic resistance (and relevant to the current paper, serotype 19A is commonly associated with antibiotic resistance). More generally, a public health analysis should consider all disease syndromes preventable by PCVs, not just IPD, and these outcomes also will be impacted by the specific serotype mix of implicated pneumococci.

We argue that a better public health approach is to optimize the benefits of pneumococcal vaccination by increasing vaccination coverage across the most vulnerable populations and for countries to assess the totality of the evidence in order to make informed decisions regarding the choice of vaccine that will provide the broadest protection across all ages. One clear decision point is assessing whether local pneumococcal epidemiology suggests disease risk will be reduced and population health improved by using a vaccine that prevents serotype 19A.

Declaration of interest

All authors are Pfizer employees and hold stocks/stock options in Pfizer Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgments

Scott Vuocolo PhD (Pfizer) provided formatting support to the authors.

Additional information

Funding

The letter was funded by Pfizer vaccines medical.