To Boost or Not to Boost? Lessons from the Australian Pneumococcal Conjugate Vaccination Program

(See the Major Article by Meder et al on pages 2607–15.)

Over the last 2 decades, the introduction of pneumococcal conjugate vaccines (PCVs) has led to remarkable reductions in the incidence in pneumococcal infections and has transformed the epidemiology of these diseases. PCV coverage continues to expand in the world, and by many accounts PCVs represent a modern-day public health success. Nevertheless, although the benefits of PCVs have been consistently demonstrated, the magnitude of the population-based impact of PCVs is difficult to quantify accurately, and impact estimates have varied among countries.

Because the real-world demonstration of PCV impact is mainly derived from surveillance systems that are unique to each country, differences in surveillance methods are often cited as an explanation for the observed differences. Other factors have also been postulated, including the vaccination schedule, the speed of vaccination and the vaccination coverage, variations in carriage and transmission patterns of Streptococcus pneumoniae, the distribution of risk factors for disease in the population, and the interaction between serotypes’ distribution and invasiveness in each setting, among others [1]. Although none of these factors is likely to be the sole explanation for the observed variability in the impact of PCVs, some factors have been more studied than others. Of note, several reports on PCV impact come from countries with vaccination schedules that include booster doses. In this issue of Clinical Infectious Diseases, Meder and colleagues describe the impact of the Australian infant PCV vaccination program with a 3 + 0 schedule and without a booster dose, an important distinction from programs from other high-income countries.

The US infant pneumococcal vaccination program (using a 3 + 1 schedule [2, 4, 6, and a booster dose at 12 months]) has led to substantial reductions in invasive pneumococcal disease (IPD) overall, vaccine-serotype IPD, and noninvasive pneumococcal diseases among both vaccinated and unvaccinated subjects, without evidence of substantial serotype replacement disease [2]. In the United Kingdom, widespread use of PCVs (2 + 1 schedule [2, 4, and a booster dose at 12 months]) has led to similar reductions in vaccine-type IPD in both children and adults, as well as pneumococcal pneumonia, all-cause pneumonia, and empyema primarily in children [3–5]. Distinct from the US experience, sizable increases in serotype replacement disease have been documented, although the rates of pneumococcal diseases in the post-PCV introduction years remain lower than the rates from the pre-PCV years [4, 6].

Using national Australian surveillance data, Meder and colleagues examined the changes in IPD and noninvasive pneumococcal pneumonia among indigenous and nonindigenous populations. The study used laboratory surveillance data for detection of IPD and available hospitalization records for the study of pneumococcal pneumonia (identified through recorded coded ICD diagnoses) from 2002 through 2016. In Australia, funded PCV vaccination programs started in 2001 for aboriginal and high-risk children, while vaccination of all children with PCV7 and adults age ≥65 years with pneumococcal polysaccharide vaccine (PPV23) started in 2005; PCV13 replaced PCV7 in 2011. Three-dose vaccination coverage for both indigenous and nonindigenous children was around 90% since 2005. The incidence of pneumococcal diseases was estimated and compared among periods: preuniversal PCV7 (2002–2004), early-PCV7 (2005–2007), pre-PCV13 (2008–mid 2011), and post-PCV13 (mid 2011–2016).

The article described major declines in IPD incidence and also declines in pneumococcal pneumonia in all age groups following the introduction of PCV7. More modest reductions in IPD followed the switching to PCV13 in young children but not in older age groups. For pneumococcal pneumonia the incidence was declining before introduction of PCV7 in 2005, reached a low level in 2006, started increasing among subjects age ≥5 years since 2007, and continued increasing in spite of introduction of PCV13 in 2011. The article also provides comparisons between indigenous and nonindigenous Australians and showed modest impact of vaccination among indigenous populations on IPD and pneumococcal pneumonia. The declines in IPD and pneumococcal pneumonia among young children are consistent with other experiences with PCV vaccination programs. However, as the investigators discussed, the comparisons by aggregated periods of time make it somewhat difficult to disentangle vaccination impact from other underlying trends.

The current study extends previous reports that had described a positive but limited impact of PCVs on IPD relative to reductions observed in countries with a booster dose in the vaccination schedule, increases in number of laboratory-confirmed vaccination failures—especially after switching to PCV13, and limited indirect protection with the 3 + 0 schedule in Australia [7–9]. A previous examination of a 9-year period that also included both PCV7 and PCV13 introduction in Australia demonstrated that, although IPD rates declined after PCV7 introduction for all age groups with a major reduction in PCV7 serotypes, the reductions associated with PCV13 were more modest than in other countries, mainly due to a limited impact on serotypes exclusively covered by PCV13 and due to increases in disease due to nonvaccine serotypes [7]. Further studies highlighted concerns regarding the waning protection of PCV13 without a booster dose, including an almost 8-fold increase in the number of vaccine failure IPD cases (vaccine-serotype IPD among vaccinated children) from 2008 to 2017, a period with widespread use of PCVs [8, 9]. The observation that almost all vaccine failures occurred in children >12 months of age suggested that a booster dose may have been helpful. These observations contributed to a growing body of evidence that ultimately led to the decision to switch the Australian PCV vaccination program into a 2 + 1 schedule in 2018 [9].

The decision to switch the PCV schedule aligns the Australian vaccination program with those from other high-income countries. Yet, a systematic review of existing evidence on indirect protection derived from pneumococcal conjugate vaccines did not conclusively determine the role of a booster dose on the prevention of IPD and noninvasive pneumococcal diseases [10]. Currently, the World Health Organization (WHO) recommends a 3-dose schedule, either as a 2 + 1 or as 3 + 0, suggesting consideration of vaccination timeliness and expected coverage during the schedule selection. The WHO considers that the 2 + 1 schedule has potential benefits over the 3 + 0 schedule as higher antibody levels are induced in the second year of life, which is thought to be important in maintaining indirect protection as toddlers are considered the major source of pneumococcal transmission in the community [11, 12]. Because additional information is necessary to support a schedule preference, a close examination of the Australian and other experiences from settings without booster doses is warranted.

Is the added booster dose going to remove activity of all vaccine serotypes in Australia? It is unclear for now, but findings from other countries that have booster doses in their PCV schedule have consistently reported persistent residual activity of some serotypes currently covered by the vaccines, especially in older unvaccinated groups. Although the apparent limited protection for children that did not receive a booster dose was a major concern leading to the switch to the 2 + 1 schedule, an important question remains as to why vaccine failures appeared to intensify after switching to PCV13 in Australia. Although potential changes in surveillance may have played a role, available evidence indicates that PCV7 on a 3 + 0 schedule was equally protective against carriage, noninvasive, and invasive pneumococcal disease as the 2 + 1 and 3 + 1 schedule in spite of lower observed immunogenicity associated with the lack of a booster dose [13]. The same amount of evidence is not apparent for the use of a 3 + 0 schedule with PCV13. Even though studies have demonstrated that the immunological response to PCV13 is noninferior to PCV7, some but not all studies have described numerically lower immune responses for some serotypes [14–16]. Although the biological relevance of those observations is unclear, some studies have also documented improved immune responses against serotype 19F among recipients of PCV13 compared with PCV7 [15]. Interestingly, 19F was commonly detected among vaccine failure IPD cases detected in Australia, especially during the PCV13 period [9]. Clearly, additional detailed studies, as those conducted in Australia, would be useful in other settings to continue monitoring the occurrence of pneumococcal disease in the post-PCV13 era.

The observations described by Meder et al and others [7–9] provide detailed assessments of the impact of the Australian PCV vaccination program. Similar studies will be necessary to assess the impact of the new vaccination schedule in Australia. As other existing programs are undergoing adjustments, similar evaluations will be warranted. After demonstration that post-booster responses were similar in children receiving either a 2 + 1 or 1 + 1 PCV schedule, the transition from a 2 + 1 to a 1 + 1 schedule in the United Kingdom is underway [17]. Maintaining a booster dose seems very important for PCV vaccination programs, and it is considered an important driver of indirect protection [11]. As the US Advisory Committee on Immunization Practices recently discontinued their recommendation for routine vaccination of all adults age ≥65 years with PCV, the importance of the children’s PCV vaccination program cannot be overemphasized. Effective PCV vaccination programs for children are vital to continue providing protection against pneumococcal diseases to both vaccinated and unvaccinated groups.

Notes

Financial support. This work was supported in part by a grant from the National Institutes of Health and the National Institute on Aging (5R01AG043471) and a PhRMA Foundation Postdoctoral Fellowship Award in Health Outcomes.

Potential conflicts of interest. A. D. W. has no potential conflicts to disclose. C. G. G. has received consulting fees from Pfizer, Sanofi, and Merck and received research support from Sanofi-Pasteur, Campbell Alliance, the Centers for Disease Control and Prevention, the National Institutes of Health, the Food and Drug Administration, and the Agency for Healthcare Research and Quality. The authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.