Impact of pneumococcal conjugate vaccines on nasopharyngeal carriage and invasive disease among unvaccinated people: Review of evidence on indirect effects

Vaccine 32 (2014) 133–145

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Impact of pneumococcal conjugate vaccines on nasopharyngeal carriage and invasive disease among unvaccinated people: Review of evidence on indirect effects Stephanie M. Davis a,1 , Maria Deloria-Knoll a,1 , Hilina T. Kassa a,1 , Katherine L. O’Brien a,b,∗,1 a b

International Vaccine Access Center, Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, United States Center for American Indian Health, Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, United States

a r t i c l e

i n f o

Article history: Received 7 January 2013 Received in revised form 26 April 2013 Accepted 1 May 2013 Available online 16 May 2013 Keywords: Pneumococcus Pneumococcal conjugate vaccine Indirect effect Nasopharyngeal colonization Nasopharyngeal carriage

a b s t r a c t Background: Invasive disease due to Streptococcus pneumoniae remains an important worldwide cause of morbidity and mortality, particularly in young children and the elderly. The development and use of pneumococcal conjugate vaccines (PCVs) have had a dramatic impact on rates of vaccine-type invasive pneumococcal disease (IPD) not only in the pediatric population targeted for vaccination but in nonvaccinated age-groups as well. This indirect effect is directly mediated by a reduction of vaccine-type nasopharyngeal carriage and thus transmission by vaccinated children. Current PCV licensing procedures do not take into consideration nasopharyngeal carriage impact, and thus the indirect effect. This review summarizes the evidence for the indirect effect of PCV on vaccine-type disease and its correlation with changes in carriage among unvaccinated populations, to assess the basis for inclusion of carriage in the PCV licensing process. Methods: Randomized controlled trials, surveillance and other observational studies published between 1994 and 2013 were systematically identified from global, regional and review databases and conference abstracts. We included as primary evidence, studies in non-vaccinated groups addressing changes in both vaccine-type IPD and carriage between pre- and post-PCV introduction periods; studies missing one of these four components were included as supporting rather than primary evidence. Results: We identified studies from 14 countries, nearly all developed countries. Vaccine-type IPD and carriage in non-targeted populations consistently decreased after PCV introduction, with the magnitude of decrease growing over time. Where IPD and carriage were observed in the same population, VTdecreases occurred contemporaneously. These relationships held true across age-groups and between indigenous and non-indigenous populations in the US and Australia. Conclusions: Indirect PCV impact on VT-IPD and VT-carriage has been significant. Impact on carriage should be considered for inclusion in the PCV licensure process as a predictor of indirect effects. © 2013 Published by Elsevier Ltd.

1. Introduction Streptococcus pneumoniae caused over 500,000 estimated deaths among children under 5 years of age globally in 2008. [1] Adults, primarily the elderly and immunosuppressed, also suffer a high burden of mortality and morbidity from this pathogen [2]. In all age-groups there is a disproportionate burden of disease among those who live in the developing world or have limited access to treatment [3].

∗ Corresponding author at: 621 N. Washington Street, Baltimore, MD 21205, United States. Tel.: +1 410 955 6931; fax: +1 410 955 2010. E-mail address: [email protected] (K.L. O’Brien). 1 For the Pneumococcal Carriage Group (PneumoCarr) and the PCV Dosing Landscape Project. 0264-410X/$ – see front matter © 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.vaccine.2013.05.005

In 2000 the first pneumococcal conjugate vaccine (PCV) was licensed in the United States. It included the seven most common serotypes causing invasive pneumococcal disease (IPD) among young children in North America [4]. Unlike pure polysaccharide vaccines that generate a T cell-independent, antibody-mediated response, conjugate vaccines engage T-cell-mediated immunity, stimulating serotype-specific antibody production and immunologic memory, providing protection beginning in infancy against disease from included serotypes. The basis for licensing the first PCV product was clinical efficacy against vaccine-serotype (VT) IPD demonstrated through randomized, double-blind, clinical trials of infants [5,6]. Experience in the prior decade with Haemophilus influenzae type b (Hib) conjugate vaccine demonstrated decreased Hib oropharyngeal and nasopharyngeal (NP) carriage in vaccinated children, reducing transmission to and disease in unvaccinated children; this is termed the indirect

134

S.M. Davis et al. / Vaccine 32 (2014) 133–145

or herd effect. Because of the Hib vaccine experience, early PCV studies evaluated the impact on pneumococcal NP carriage as an indicator of the potential for indirect protection. Numerous studies showed that PCV reduces NP acquisition of VT carriage among vaccinated subjects [7]. However, PCV also increases the colonization prevalence of non-vaccine serotypes (NVTs) – a phenomenon termed serotype replacement – leaving overall pneumococcal carriage prevalence virtually unchanged. PCV introduction into the routine pediatric immunization schedule in the United States and other countries has resulted in near-elimination of VT-IPD not only in infants (the age-group targeted for vaccination), but also in the unimmunized general population [8]. This indirect protection is a critical component of the vaccine’s public health impact. In the United States, it accounted for 69% of all IPD cases prevented in the first three years of licensure [9] and a 44–63% absolute decrease in pneumococcal pneumonia admissions in adults [10]. PCVs have now been incorporated into routine childhood immunization in 96 countries. Another 51 countries, many in the developing world, plan to introduce PCV in the coming years [11]. With demand growing, multiple manufacturers are developing PCV products; licensing authorities have had to determine what data should support such licensure and be required for post-licensure monitoring. Disease endpoint trials are now difficult or impossible to conduct because of ethical considerations in placebo-control comparisons and sample size requirements in head-to-head trials. Licensure approaches are therefore anchored on correlations of immunogenicity to IPD protection established in the randomized controlled trials, and immunogenicity non-inferiority measures in new PCV products [12]. Although this approach has a strong scientific basis and is accepted by the European Medicines Evaluation Agency, the United States Food and Drug Administration, and the World Health Organization (WHO), it lacks a crucial component: impact of pneumococcal vaccines on NP carriage among both the vaccinated and unvaccinated, and consequent effects on disease among the unvaccinated as well as the fully or partially vaccinated. NP effects may also prove an essential component of the licensing approach for novel non-polysaccharide pneumococcal vaccines such as those based on pneumococcal proteins. Not only do vaccine products merit consideration from this perspective of impact on carriage, so do vaccine schedules; the number of primary-series doses and addition of a booster dose may affect the magnitude of the indirect effect. We posited the causal chain in the indirect effect paradigm as follows (Fig. 1): 1. PCV decreases VT-carriage prevalence and density in vaccinated individuals. Reduction in prevalence is achieved by reductions in acquisition rates and density, rather than reductions in duration of VT carriage [13–15].

PCV vaccinaon of children

Reduced VT carriage in vaccinated

Reduced VT disease in vaccinated

Reduced VT carriage in unvaccinated

Reduced VT disease in unvaccinated

Fig. 1. Posited causal chain for the relationship between PCV vaccination and VT-IPD in the unvaccinated.

2. These changes result in decreased VT-carriage transmission to, and carriage among, unvaccinated (and vaccinated) populations. 3. Because individual NP carriage acquisition is the prerequisite for pneumococcal disease, reductions in VT-carriage among vaccinated and unvaccinated age groups cause a reduction in VT-IPD in these same populations. Evidence for the first link in this chain and for individual carriage as a precondition for pneumococcal disease is addressed elsewhere [16]. Here we evaluate the evidence for the links between PCV use and reductions in both VT-carriage and VT-IPD in non-target agegroups. The most compelling evidence for this link is from studies (community-randomized trials or pre- and post-PCV observational studies) simultaneously examining rates of VT-carriage and VTIPD in non-targeted groups, with and without PCV. Also relevant are studies examining PCV-associated changes in IPD or carriage alone. Others that provide secondary supporting evidence for the validity of the causal chain include studies comparing VT-IPD or NP carriage rates in non-targeted age-groups in early vs. mature post-introduction periods (time-series analyses); those comparing these rates pre- and post-introduction in populations which are predominantly non-targeted but include some targeted individuals (“mixed” populations); and those which compare pre- and postintroduction rates of all-type (AT) IPD in non-target age-groups without distinguishing VT from NVT disease. We performed a comprehensive review of studies meeting each of these descriptions to assess the evidence for the importance of NP carriage as a component of licensure of new pediatric pneumococcal vaccine products. 2. Methods A literature review through 2005 of the PCV indirect effect on IPD has been published. [17] We performed a comprehensive literature search for the PCV Dosing Landscape Project that identified PCV observational and interventional studies with respect to immunogenicity, IPD, pneumonia and NP carriage that updated the evidence through September 2010 and added changes in carriage [18]. A subsequent literature search was performed in January 2013 to identify articles with primary evidence published after the PCV Dosing Landscape Project search; these results are reported separately from the main analyses. Articles identified by double-abstract screening that reported data on NP carriage and IPD in non-targeted age-groups were included. Review articles and book chapters were reviewed for additional citations. Appendix B.1 describes the literature review methodology. Primary evidence: Articles were included as primary evidence if they reported both pre- and post-PCV introduction periods, distinguished VT from NVT isolates, and provided results on non-targeted age-groups. Supporting evidence: Papers were considered for supporting evidence if they reported on a population, age range or year not included in the primary evidence. The following hierarchy based on descending relevance was used: 1. Data comparing early vs. late post-introduction (rather than pre vs. post-introduction) periods. 2. Data on all-type (AT) (rather than VT) IPD and carriage. Data on mixed targeted and non-targeted (rather than pure nontargeted) age-groups. This includes settings with catch-up schedules (see Appendix B.1 for the variant abstraction technique used). We abstracted the PCV product and schedule, contemporaneous

S.M. Davis et al. / Vaccine 32 (2014) 133–145

135

vaccine coverage, age range of non-targeted population, VT-IPD case counts, incidences or proportions, and VT-carriage numbers and proportions. IPD was defined as isolation of S. pneumoniae from normally sterile body sites, or adult non-bacteremic pneumonia cases diagnosed by urinary pneumococcal antigen.

Quality assessment: Articles were graded using the Child Health Epidemiology Research Group modification of the GRADE criteria [25]. This approach evaluates the evidential quality of each article and then the strength of the total body of evidence.

2.1. Data analysis

Primary evidence was found in 46 studies, and supporting evidence in 57 (Fig. 2), representing 13 countries, and 33 populations. Appendix B.2 describes excluded data points. Virtually all primary IPD and carriage data came from developed countries (Fig. 3). Primary IPD data points were identified for 12 distinct populations, in nine countries, from North America, Europe, and Oceania; primary carriage data points were identified for five populations, in five countries, from five regions. IPD was defined using only blood or only CSF specimens in three studies [26–28], urine antigen (for non-bacteremic pneumococcal pneumonia cases) in one study [29], and pneumococcal-specific ICD codes in one study [10]; one study had an unspecified diagnostic standard. [30]. All studies evaluated PCV7 except two PCV9 carriage studies [31,32].

A ‘data point’ was defined as a pre- or post-introduction prevalence in a single year, age group, and population. A ‘data set’ was defined as two data points, separated in time, from the same age group and population, typically one pre- and one post- introduction. Where possible, the ‘pre’ period was before PCV licensing in the country, excluding the year licensed unless that year’s pre-data were drawn only from months prior to introduction (Appendix B.1); the ‘post’ period began no earlier than the year following introduction. Year of introduction was based on a compilation of data from WHO [19] and VIMS [20] databases which identified the year in which PCV was widely adopted on a national or relevant regional scale. In the few cases with significant lag time between national licensure and wide adoption, the breakpoint identified by the author was used (low-coverage vs. high-coverage, or prelicensure vs. post-licensure.) Percentage change in outcome measures was calculated by comparing the most recent pre-introduction data available to each available post-introduction time point. For data presented as incidence rates and case counts, percentage change was calculated as (pre-introduction – post-introduction)/pre-introduction × 100%, where negative values for percentage change denote an increase. If the study outcome was the proportion VT of all IPD cases, percentage change was transformed into a comparable measure based on incidence rates and case counts as follows: Percentage change = [1 − ((%VT IPD post) × (%NVT IPD pre))/(%VT IPD pre) × (%NVT IPD post)] × 100%. Data were stratified by elapsed years since introduction to assess trends with time, and by age group (<5, 5 to <18, 18 to <50, 50 to <65, ≥65 years) to assess differential effects across age categories. Points not fitting within a single age stratum with minimal overlap were classified based on the oldest stratum included. Where a data point represented multiple post-introduction years (i.e., “2001–2003”), the midpoint was used to calculate the number of years since PCV introduction. Where possible, data were also stratified into populations receiving booster doses and those without, and indigenous versus general populations. Effects of different primary dose schedules are addressed elsewhere [21–24]. When both IPD and carriage were available, we compared their percentage changes to assess their relationship. When both VT-IPD and PCV coverage levels in the community over time were available, we evaluated the relationship between PCV uptake and VT-IPD impact. Countries that implemented a catch-up schedule in those <2 or <5 years were identified; since catch-up coverage is generally less than complete, we did not further distinguish the magnitude of indirect effects by use of catch-up but considered these mixed populations. Due to heterogeneity of reporting formats and because weighted means could not be calculated for studies not reporting population size, the data were not considered appropriate for meta-analysis. Therefore, we estimated median percent change in outcome parameters from pre-introduction. Because indirect effects in mixed groups of targeted and nontargeted age-groups are difficult to separate from direct effects among targeted children within them, we compared single-dose coverage rates (the highest possible measure of coverage), where known, with rates of decrease in IPD in these groups. Where the latter exceed the former, an indirect component is suggested.

3. Results

3.1. Simultaneous carriage and IPD – primary evidence Both NP carriage and IPD changes following PCV introduction were available in four non-target groups: three indigenous population groups (Alaska Natives, American Indians and Australian aboriginals) and one general population group (Portugal) (Table 1). In general, percentage decreases in VT-IPD rates were within 20 percentage points of contemporaneous decreases in VT carriage rates, with decreases in VT-IPD usually but not always larger. In the only case of significant divergence (78% decrease in VT-carriage vs. 19% in VT-IPD), PCV introduction was confined to the private market, the NP and IPD data were not from contemporaneous timeperiods, and different age-groups were represented (the target age-group vs. all residents) [33,34]. The major United States IPD surveillance studies, Active Bacterial Core Surveillance (ABCs) and Northern California Kaiser Permanente Database, do not include carriage surveillance. However, comparison can be drawn between the 64–77% decline in VT-IPD from pre-introduction to 4 years post-introduction among adults in ABCs areas and the 61% relative decrease (from 36% 1 year post-introduction to 14% 4 years post-introduction) observed in cross-sectional VT carriage prevalence among all pneumococcal strains in children under 7 years of age presenting for clinic visits in Massachusetts [35]. Only the Alaska Native and Australian Aboriginal populations had high (≥50%) pre-introduction VT carriage (Appendix B.3, Table 5; data from older children and teenagers). Therefore, it remains unclear whether the relationship between impact on NP carriage relative to that for VT-IPD varies with preexisting carriage burden. 3.2. IPD-only – primary evidence Primary evidence included 38 articles representing 9 countries and 26 populations (some overlapping), including indigenous populations, HIV and AIDS patients, and the general population. PCV introduction was nearly invariably followed by sharp reductions in VT-IPD rates in non-targeted populations, including infants too young to be immunized [36] (Appendix B.3, Table 1). The median proportion decrease in VT-IPD incidence among unimmunized age-groups increased with number of years post routine PCV introduction (Table 2). Of 56 age-specific data points,

136

S.M. Davis et al. / Vaccine 32 (2014) 133–145

Fig. 2. IPD and NP carriage article selection process.

53 reported decreases in VT-IPD incidence. All age-groups experienced significant indirect benefit, with many data points showing declines in VT-IPD below 50% and near elimination for those with the longest follow-up (Fig. 4). Median percentage decrease in VT-IPD was 57% (interquartile range [IQR]: 40–77%) for the general population, 67% (IQR: 40–85%) for aboriginal populations, and 30% (IQR: 13–46%) for HIV-positive populations (data not shown). Plateaus in values should not be interpreted to mean that within a population this plateau is observed since values reflect data from varying settings and countries. PCV vaccination coverage among targeted age-groups was reported in heterogeneous formats across the various publications,

limiting summary correlations between VT-IPD changes among non-targeted age-groups and coverage (Table 3) although these seemed to correlate over time. When coverage rates were high, evidence for indirect impact was consistent; it was mixed with low coverage rates but suggestive, starting at 3-dose coverage among 19–35-month-olds as low as 40%. If PCV target-aged children were the only significant pneumococcal carriers in communities, rates of VT-IPD in all age-groups might fall proportionate to some function of coverage soon after introduction. Instead, decreases in VT-IPD in non-target groups exceed contemporaneous 3-dose vaccine coverage rates in their communities (Table 3). In the US ABCs and Navajo populations where vaccine has been used the longest albeit with imperfect coverage, VT-IPD among non-target groups has been virtually eliminated in the 5–10 years following introduction. Six data sets (all from Australia) evaluated a primary series schedule without a PCV booster dose; the median decrease of VTIPD among non-target groups was 60% (IQR: 50–67%). The median decrease in VT-IPD in countries using a PCV booster dose was 62% (IQR: 40–78%) [37–45]. 3.3. IPD-only – supporting data

Fig. 3. Locations of populations reporting primary IPD and NP carriage data.

Appendix B.4 includes a full discussion of supporting data. Briefly, comparison of early and late post-introduction VT-IPD rates shows an impact similar to that seen over comparable time periods in the primary evidence (Appendix B.3, Table 2). Evidence on indirect impact in low-coverage (<70%) settings is mixed, with significant impact seen in some populations and not others. Data on indirect effect of PCV on AT–IPD showed a trend toward increasing impact with time (median decrease: 33%; IQR: 7–42%),

S.M. Davis et al. / Vaccine 32 (2014) 133–145

137

Table 1 Data addressing contemporaneous changes in VT-carriage and VT-IPD rates in identical or similar non-target groups between pre- and post-PCV introduction periods (References are found in Appendix B.6). Carriage

IPD

Population and age in years

Years

Reference group for decrease

% decrease

Population and age in years

Years

Decrease type

% decrease

95% Alaska Native (AN) children 2–4b in 8 Alaska villages [B9]

2000 2001 2002 2003

Carriers

(ref) 56 75 87

AN 2–4a , b under Arctic Investigations surveillance [B9,B13]

1995–2000, 2001–2003

Rate

(ref) 100

2000 2001 2002 2003

Carriers

(ref) 40 60 64

AN 5–17a , b under Arctic Investigations surveillance [B9,B13]

95% AN children 5–17 in 8 Alaska villages [B9] 95% AN in 8 Alaska villagesa [B38,B68] 18–24

Rate 1995–2000, 2004–2006

AN under Arctic Investigations surveillance 18–44 [B9,B13]

(ref) 100

1995–2000, 2001–2003 2004–2006

Rate

(ref) 85 100

1995–2000, 2001–2003

Rate

(ref) 7

1998–2000 2001 2002 2003 2004

Study pop

1998–2000 2001 2002 2003 2004

Study pop

(ref) 14 −25 (increase) 65 66

Study pop

(ref) −39 (increase) 10 7 52

Study pop

71

1998–2000 2001 2002 2003 2004

Study pop

(ref) 2 68 55 62

AN ≥45 under Arctic Investigations surveillance [B9,B13]

Australian aboriginal children 4–13a , Northern Territory [B69]

2000–2001, 2002 + 2004

Study pop

(ref) 45

Australian aboriginal 5–14a , Northern Queensland [B4,B5]

1999–2001, 2002–2004 2005–2007

Case No. Rate

(ref) 66 50

Australian aboriginal parents of young infants, Northern territory [B69]

1996–1997 and 1999–2001, 2002 + 2004

Study pop

(ref) 45

Australian aboriginal ≥15, Northern Queensland [B5]

1999–2001, 2005–2007

Rate

(ref)

Rate

75

Rate

(ref) 29

25–34

35–44

1998–2000 2001 2002 2003 2004 1998–2000, 2004–5

≥18

95% AN >45 in 8 Alaska villages [B68]

(ref) 23 53 72 82

≥18 [B13,B38]

Rate 1995–2000, 2004–2006

Rate

1995–2000, 2004–2005

1995–2000, 2001–2003

(ref) 63

Rate

(ref) 10

Rate 1995–2000, 2004–2006

Australian aboriginal 15–49, Northern Queensland [B4] Australian aboriginal 50+, Northern Queensland [B4]

(ref) 67

1999–2001, 2002–2004

Australia Northern Territory 50+ (may be mixed aboriginal/not) [B31]

(ref) 70

Rate (ref) 60 (ref) 82

1999–2001, 2002–2004

1994–2000, 2002–2005

138

S.M. Davis et al. / Vaccine 32 (2014) 133–145

Table 1 (Continued ) Carriage

IPD

Population and age in years

Years

Reference group for decrease

% decrease

Population and age in years

Years

Decrease type

% decrease

Apache and Navajo <5y household members of vaccinated children [B70]

Initially intervention (continued through 2002)

Study pop

37

Apache and Navajo 2 to <5a [B2]

1996 2001

Case No. Case No.

ref–RCT 33

Navajo 2 to <5a [B11]

1995–1997, 2001–2005

Rate

(ref) 87

1995–1997, 2007–2009 Apache and Navajo 5 to <18 years household members of vaccinated children [B70]

Initially intervention (continued through 2002)

Study pop

6

Apache and Navajo 5 to <18 [B2]

1996 2001

Navajo 5 to <18a [B3,B11,B71]

1995–1997, 2001–2005

(ref) 100 Case No. Case No.

1995–1997, 2001–2003 2004–2006

Rate Rate

Initially intervention (continued through 2002)

Study pop

54

Study pop

41

Apache and Navajo 18k to <40 [B2]

1996 2001

Apache and Navajo 40 to <65 [B2]

1996 2001

Apache 18 to <50 [B15]

1997–2000, 2001–2004

Case No.

Navajo 40–64 [B3,B11,B19] Apache and Navajo 65+ [B19] Navajo 65+ [B3,B11,B19]

Case No. Rate Rate

1995–1997, 2001–2005

Rate

1995–1997, 2001–2003 2004–2006

Rate

1995–1997, 2007–2009

Rate

Case No.

Rate 1995–1997, 2001–2005 1995–1997, 2001–2003 2004–2006

Case No.

Rate

1995–1997, 2007–2009

a b

Approximated from graph. Population contains some target-aged subjects (is only partially non-target).

(ref) 28 (ref) 94 (ref) 10 54

(ref) −5 (increase) 81

2004–2006

1999–2001, 2002–2004

(ref) 0 40

(ref) 48

1995–1997, 2001–2003

Portuguese residents of 4 of 5 mainland regions 1–5 [B7]

(ref) 40

(ref) 16

1995–1997, 2001–2005

(ref) 78

(ref) −600 (increase)

(ref) 56

1996, 2001

Study pop

(ref) 63

Rate

1995–1997, 2007–2009

2001 2006

(ref) 41

Case No.

Navajo 18 to <40 [B11,B19]

Portuguese (Lisbon and Oeiras) daycare attendees 4 months to 6 years [B72]

(ref) 0 0 (ref) 66

1995–1997, 2007–2009 Apache and Navajo ≥18 household members of immunized children [B70]

(ref) 0 (no cases) (ref) 40

(ref) 90 Case No.

(ref) 19

Case No. Portuguese residents of 4 of 5 mainland regions 6–18 [B7]

(ref) −60 (increase)

S.M. Davis et al. / Vaccine 32 (2014) 133–145

139

Table 2 Median percentage decrease in VT-IPD rate in non-targeted groups, by number of years after PCV introduction. Years after introduction

Number of data points

Median decrease

1 2 3 4 5 6 8 11

10 11 7 15 5 1 8 4

21 59 54 70 54 48 87.5 78

though with lower overall impact compared to that on VT-IPD (Appendix B.3, Table 3). This impact on AT-IPD was observed in all non-target age-groups (Fig. 5) and is also noted in pneumococcal pneumonia [10,29]. Data from mixed target and non-target groups show a greater decrease in VT-IPD rates than that in pure non-targeted groups, reflecting a mix of direct and indirect effect (Appendix B.3, Table 4). However, studies with 1-dose coverage data suggest a vaccine impact on VT-IPD that cannot be entirely accounted for by direct effect.

a

d

Interquartile range (%) 7–40 46–69 33–75 40–78 50–69 – 76.5–92.5 47–92

3.4. Carriage-only – primary evidence Data were available for six unique populations: Australian aboriginals, Alaska Natives, American Indians, Gambians, Israelis and Portuguese (Appendix B.3, Table 5). Studies in children were primarily RCTs; those in adults were primarily observational. The median decrease in VT-carriage prevalence (among either the study sample or, rarely, the subset who were carriers of any pneumococcal strain) was 77% (IQR 64–80%). Data points did not span a sufficient time range to evaluate time-related trends. The majority of carriage data is drawn from high-risk populations.

b

e

c

f

Fig. 4. Percent decrease of VT-IPD, by time since vaccine introduction, by age group. 1–38. a = <5 years, b = 5 to <18 years, c = 18 to <50 years, d = 50– < 65 years, e = ≥65 years, f = all. indicates points representing broader age ranges up to 5–64 years in the 50 to <65 age category. Where subpopulations or overlapping populations had data on years not included in the larger population, data points were those with age ranges fitting closely rather than broadly into their age category, with specific years rather than periods, and based on data from tables rather than extrapolated from graphs. Data points with particularly broad age ranges – those which had a lower age bound <40 and no upper age bound – are displayed only in the all-ages graph; overlapping data points with more tightly-fitted age groups are displayed only in the appropriate age-specific graph.

140

S.M. Davis et al. / Vaccine 32 (2014) 133–145

Table 3 Data presenting PCV vaccination coverage in targeted age-groups and VT-IPD data in non-targeted age-groups (References are found in Appendix B.6). Population

Navajo children [B3]

Coverage population

Non-targeted population

Year of introduction

Age at assessment

Coverage (year)

Population

Age (years)

% decline in IPD (end year(s))

1997 (selected communities); 2000 (national introduction)

19–35 months

≥3 doses 55–65% (2002) 80–95% (2003) 87–100% (2004) 85–100% (2005)

Navajo [B3]

5–<18a

0 (2001–2003) 0 (2004–2006) (NS) 66 (2007–2009)

18–<40

0 (2001–2003) 40 (2004–2006) 28 (2007–2009) 10 (2001–2003) 54 (2004–2006) 94 (2007–2009) –5b (2001–2003) 81 (2004–2006) 90 (2007–2009)

40–<65

65+

White Mountain Apache [B10]

Alaska Native children [B13]

Children in all Alberta [B6]

1997 (selected communities); 2000 (national introduction)

19–35 months

1997 (selected communities); 2000 (national introduction)

19–35 months

2002

12–24 months

≥3 doses 9% (1998) 40% (2001) 89% (2004)

White Mountain Apache [B10]

≥3 doses 88% (2003) 96% (2006)

Alaska Natives [B9,B13]

≥3 doses 91% (2006)

5–17

84 (2004–2006)

≥18

80 (2004–2006)

5–17

85 (2001–2003) 100 (2004–2006)

18–44 ≥45

73 (2001–2003) 10 (2001–2003) 70 (2004–2006)

Calgary and surrounding communities (subset of Alberta population) [B6]

5–15 16–64 64–84 ≥85

45 (2003–2007) 38 (2003–2007) 78 (2003–2007) 23 (2003–2007)

Alaska non-native [B9,B13]

5–17

–150 (2001–2003) 36 (2004–2006)

≥4 doses 84% (2005) Alaska non-Native (white) children [B9]

2000

19–35 months

≥3 doses 67.4% (2003)

18–44

≥45 Spain [B24]

2001

<5 years

≥1 dose 5% (2001) 11% (2002) 20% (2003) 30% (2004) 42% (2005) 51% (2006) 59% (2007)

Navarre, Spain residents [B24]

5–64

US white children living in 7 ABCs area [B26] US black children living in 7 ABCs area [B26] a

2000

2000

2000

15 months

19–35 months

19–35 months

≥1 dose 23%(2000) 63% (2001) 85% (2002) 85% (2003)

US Metropolitan Atlanta [B14]

≥3 doses 7% (2001) 43% (2002)c

US whites living in 7 ABCs surveillance areas [B26]

≥3 doses 6% (2001) 41% (2002)c

Population contain some target-aged subjects (is only partially non-target) b Negative values denote an increase. c Estimated from graph (may be imprecise). NS = not significant.

5 (2003–2005) 45 (2006–2007)

65+ Inpatients at Hospital de, Barcelona [B73]

18–64

65+ Largest 2 urban Atlanta counties [B14]

75 (2001–2003) 80 (2004–2006) 31 (2001–2003) 81 (2004–2006)

All adults

6 (2003–2005) 68 (2006–2007) 2 (2002–2004) 12 (2005–2007) 25 (2002–2004) 37 (2005–2007) 68–72 (2003–2004)

65+ 17 (2000–2001) 28 (2001–2002) 44 (2002–2003) 56 (2003–2004)c

US blacks living in 7 ABCs surveillance areas [B105S]

18–64

51 (2002)

65+

46 (2002)

18–64

69 (2002)

65+

47 (2002)

S.M. Davis et al. / Vaccine 32 (2014) 133–145

141

a

b

d

e

c

f

Fig. 5. Percent decrease of AT-IPD, by time since vaccine introduction, by age group 39–67. a = <5 years, b = 5 to <18 years, c = 18 to <50 years, d = 50 to <65 years, e = ≥65 years, f = all. The large-population surveillance networks provided primarily unique data but in some cases have small overlaps in coverage area.

3.5. Carriage-only – supporting evidence

3.7. Articles published After September 2010

Few additional supporting data points were identified for NP carriage. Supporting data are listed for pre- vs. post-introduction all-type NP in non-target groups and pre- vs. post-introduction VT-carriage in mixed groups in Appendix B.3, Tables 6 and 7; a discussion is provided in Appendix B.4. A relevant data point not eligible for inclusion due to publication date comes from an observational study including Native American adults shortly after PCV introduction (2001–2002) and subsequently (2006–2008), finding a relative decrease of 97.5% and an absolute reduction of 4.0% in VT-NP [46].

An additional 14 studies published after the PCV Dosing Landscape Review search met primary evidence inclusion criteria. The two articles reporting change in VT-carriage in nontarget groups noted a 30% decrease in four-year-olds and 19% in five-year-olds in daycares throughout Hong Kong one year after PCV7 introduction [47], and an 89% decrease in parents of vaccinated as compared to control children in the Netherlands [48]. Ten studies extended VT IPD follow-up of studies already included in analyses; all showed persistent decreases in VT-IPD from baseline ranging from 20% to 100%, in HIV patients in Spain [49] and in general populations in Australia [50,51] the US (ABCs) [52,53] Canada [54,55] England and Wales [56], Germany [57] and Denmark [58]. VT IPD in 5–14-year-old inpatients with community-acquired pneumonia in Montevideo, Uruguay, a population not previously addressed, decreased 22% one year after introduction [59]. The last study was a hospital case series in Australia with only one IPD case in each pre- and post-introduction period [60].

3.6. Study quality Most individual data points were categorized as low or verylow quality by GRADE criteria because nearly all data were from observational studies, and over half the primary evidence sources were further downgraded for including only highrisk populations, but few for methodological issues (Appendix B.5). While GRADE methodology categorizes observational studies as ‘low quality’, the GRADE system was designed to assess individual patient treatments, not to assess public health benefit. Furthermore, only observational, or community randomized studies can assess population-level post-introduction effects.

4. Discussion This review summarized data from 14 countries, demonstrating the breadth of PCV impact on NP carriage and IPD among age groups not targeted for vaccination. Introduction of PCV into communities is consistently followed by significant decreases in both VT-carriage

142

S.M. Davis et al. / Vaccine 32 (2014) 133–145

Table 4 Countries included in analyses in which children <5 years received PCV as part of a catch-up campaign (References are found in Appendix B.6). Country

Catch-up target age range and schedule

Australia (indigenous) [B74]

Catch-up program in June–July 2001 for indigenous children aged up to 2 years in the northern region and up to 5 years in the central region of the Northern Territory Catch-up program in 2005 only for children born 1 January 2003–31 December 2004 (inclusive): 2 doses for unvaccinated age 7–17 months1 dose for unvaccinated age ≥ 18 months 2 doses 8 weeks apart without a booster in children 12–23 months; one dose, no booster in children 24–59 months. A single dose in children 13–24 months of age Sept 2006-Sept 2007 A single dose in children 13–24 months of age Sept 2006-Sept 2007 A single dose for healthy unvaccinated children 2–4 years and 2 doses given 2 months apart for unvaccinated children 2–4 years with certain chronic conditions, except during the vaccine shortage August 2001–May 2003, where catch-up and the fourth dose were recommended deferred in healthy children.

Australia (general) [B75]

Canada [B76]

England and Wales [B77] Scotland [B78] USA79–82

and VT-IPD in these groups. This pattern argues that carriage is the mechanism for the VT-IPD change, mediating the role of vaccination in stopping transmission from young children to other age-groups. Where data on both VT-carriage and VT-IPD exist in the same groups, decreases are contemporaneous, and although their greatest magnitude is in the first few years following PCV introduction, longitudinal data generally show continued declines [61–67]. Impact is clearest at high vaccination coverage levels but visible with coverage as low as 40%. It is seen across age-groups. The supporting data suggests a similar indirect impact. In “mixed” under-5 age-groups (i.e. combining direct and indirect effects), indirect protection is visible through impact exceeding target-group vaccine coverage, albeit in some populations introduction included a catch-up schedule (Table 4). Larger impact was observed in observational studies than in RCTs, presumably because herd effects are stronger after widespread introduction than in individually randomized studies. The magnitudes of VT-carriage reductions and those of VT-IPD are not always parallel. However, even the communities with the smallest ratio of VT-IPD decline to VT-carriage decline experienced a decrease sufficient to represent a dramatic public health gain. Additionally, decrease in VT-carriage is proposed not as an ideal proxy for expected indirect impact – it does not fully measure colonization density changes which also impact IPD risk–but as one mediator in the relationship between direct impact of PCV on VT-carriage in target groups and indirect protection, and as an improvement to the current licensing process which does not consider indirect impact at all. The few studies with VT-carriage or VT-IPD impact findings inconsistent with these trends have unique attributes. Most primary data points with no decrease in VT-IPD in non-targeted populations, had small denominators and/or small pre- and post-introduction cases (n = 0–6), resulting in unstable estimates [34,61,66,68]. Similarly, the two points categorized as showing no decreases in VT-carriage among non-target age-groups came from a community-randomized controlled trial in American Indians, in which VT carriage prevalence in young infants and older siblings of vaccinated subjects decreased nearly 50% without achieving statistical significance [13].

Similarly, the few AT-IPD data points that showed increases after vaccine introduction, in Spanish, Canadian and American populations, were attributed by their authors to increases to improved surveillance [69] and to the 2005–2006 Canadian serotype 5 outbreak [70]. Other studies showed minimal increases, reflecting essentially unchanging rates [71,72] or did not meet statistical significance [26]. The 13% statistically significant increase in AT-IPD among 50–64-year-olds in Sydney was an isolated increase against a context of IPD falling in the general population and the other agegroups studied [73]. A few increases were significant and remain unexplained [74,75]. AT-IPD trends potentially reflect the extent of serotype replacement (reviewed separately [76]), but are also subject to confounding [77] by secular trends, changes in surveillance methodology, variability in viral seasonality, and antibiotic use [78]. Under-representation of developing-world settings is a limitation of this review. This is expected, as routine PCV use is in early implementation among developing countries; the degree to which similar indirect effects will occur is uncertain. Given the higher prevalence of NP carriage in children beyond the vaccine age range in many of these settings, vaccination may miss a larger proportion of the total transmitting group, especially when catchup campaigns are not used. More generally, carriage data is quite sparse. Inferences about changes in NP carriage due to PCVs are also limited by differences in pre-introduction carriage prevalence between strains and PCV products used. Serotypes 1 and 5 are rarely carried so are not amenable to carriage studies using conventional microbiologic techniques. Implementation of newer molecular lab approaches for identifying and serotyping pneumococci may reveal more carriage for these strains than appreciated to date. Impact of a PCV booster dose on disease relative to carriage also could not be assessed as only one country (Australia) without a booster dose had both IPD and NP data; no differences from the general trends were evident. Additional such data will soon be available from Kenya and The Gambia. Meta-analysis of the relationship between the major parameters of interest was not attempted due to the heterogeneity of pathogen and vaccine metrics (years vs. periods, measures of vaccination coverage, and age-group cutoffs). Sources of simultaneous longitudinal IPD and carriage data remain sparse, but results of ongoing studies in The Gambia, South Africa, Kenya, Israel, and the US (American Indian populations, Alaska and Massachusetts) will deepen our understanding of quantitative and temporal relationships. Limiting comparisons to the latest pre-introduction years limited our ability to incorporate pre-introduction temporal trends. Conversely, abstraction of only the earliest full post-introduction year for data points in those <5 years of age, to maintain a “pure” non-targeted group, resulted in exclusion of later data points when the PCV impact would be greater. Finally, we did not assess indirect effects in vaccinated children. Because direct protection from vaccination is imperfect and vaccinated children remain at some risk for disease, some component of their protection is likely due to indirect effects. This is supported by declines in all-cause pneumonia in vaccinated age groups after introduction significantly exceeding those found in pre-licensure efficacy trials [79]. Additionally, although pneumonia is by far the most common clinical syndrome associated with pneumococcal infection, most cases of pneumococcal pneumonia are not microbiologically identified and thus not represented here. However, the included pneumonia data are consistent with the relationships described. In spite of these limitations, the consistent association between PCV introduction and subsequent declines in both VT-carriage and

S.M. Davis et al. / Vaccine 32 (2014) 133–145

VT-IPD in non-target age-groups supports reduction of NP carriage and transmission as a key element in the overall public health impact of PCV, offering a unique contribution for licensing decisions for pneumococcal vaccines. Acknowledgements The authors gratefully acknowledge the work of Jennifer Loo for provision of the literature search results. This study is part of the research of the PneumoCarr Consortium funded by the Grand Challenges in Global Health Initiative which is supported by the Bill & Melinda Gates Foundation, the Foundation for the National Institutes of Health, the Wellcome Trust and the Canadian Institutes of Health Research. We gratefully acknowledge the Pneumococcal Conjugate Vaccine Dosing Landscape project, a project of the Accelerated Vaccine Initiative, Technical Assistance Consortium-Special Studies. Support for the Pneumococcal Conjugate Vaccine Dosing Landscape Project, was provided by Program for Appropriate Technology in Health (PATH) through funding from the Global Alliance for Vaccines and Immunization (GAVI). The views expressed by the authors do not necessarily reflect the views of the GAVI Alliance and/or PATH. Conflict of interest statement: KOB has had research grant support related to pneumococcus from Pfizer, and GlaxoSmithKline and has served on pneumococcal external expert committees convened by Merck, Aventis-pasteur, and GlaxoSmithKline. MDK serves on a Data and Safety Monitoring Board for Novartis for vaccines unrelated to pneumococcus. Among the PneumoCarr members and PCV Dosing Landscape Project members the following declarations were made. KPK has had research grant support from Pfizer and has served on pneumococcal external expert committees convened by Pfizer, Merck, Aventis-pasteur, and GlaxoSmithKline. RD has received grants/research support from Berna/Crucell, Wyeth/Pfizer, MSD, Protea; has been a scientific consultant for Berna/Crucell, GlaxoSmithKline, Novartis, Wyeth/Pfizer, Protea, MSD and a speaker for Berna/Crucell, GlaxoSmithKline, Wyeth/Pfizer; he is a shareholder of Protea/NASVAX. JAGS has received research grant support from GSK and travel and accommodation support to attend a meeting convened by Merck. SAM has had research grant support from GlaxoSmithKline anmd Pfizer, and has served on pneumococcal external committees convened by Pfizer, MERCK and GlaxoSmithKline. DG has received honoraria for participation in external expert advisory committees on pneumococcal vaccines convened by Pfizer, GSK, Sanofi Pasteur and Merck. His laboratory performs contract research for Merck, Sanofi Pasteur and GSK. MGL has served as speaker in several GSK conferences and as member of two GSK advisory board meetings. HN has served on pneumococcal vaccination external expert committees convened by GlaxoSmithKline, Pfizer, and sanofi-pasteur. Other authors report no potential conflicts of interest. Appendix A. A.1. PCV dosing landscape members Laura Conklin: Centers for Disease Control and Prevention, Atlanta, GA; Katherine Fleming-Dutra: Centers for Disease Control and Prevention, Atlanta, GA; David Goldblatt: Institute of Child Health, University College London, UK; Jennifer Loo: Centers for Disease Control and Prevention, Atlanta, GA; Daniel Park: International Vaccine Access Center, Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; Cynthia Whitney: Centers for Disease Control and Prevention, Atlanta, GA.

143

A.2. PneumoCarr Consortium Primary Consortium Partners Martin Antonio: Medical Research Council Laboratories, Fajara, The Gambia; Kari Auranen: Finnish National Public Health Institute, Helsinki, Finland; Ron Dagan: Soroka University Medical Center and the Faculty of Health Sciences, Ben-Gurion University of the Negev; David Goldblatt: Institute of Child Health, University College London; Helena Kayhty: Finnish National Public Health Institute, Helsinki, Finland; Keith P. Klugman: Rollins School of Public Health, Emory University, Atlanta; Marilla Lucero: Research Institute of Tropical Medicine, Philippines; Shabir Madhi: Respiratory and Meningeal Pathogens Research unit, University of Witwatersrand; Kim Mulholland: University of Melbourne, Australia; Hannah Nohynek: Finnish National Public Health Institute, Helsinki, Finland; J. Anthony G. Scott: Oxford University, KEMRI Wellcome Trust Unit, Kilifi, Kenya. Appendix B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.vaccine. 2013.05.005. References [1] Estimated Hib and pneumococcal deaths for children under 5 years of age, 2008. World Health Organization; 2012 March [accessed online 02.06.12] at http://www.who.int/immunization monitoring/burden/Pneumo hib estimates/en/index.html [2] Wroe PC, Finkelstein JA, Ray GT, Linder JA, Johnson KM, Rifas-Shiman S, et al. Aging population and future burden of pneumococcal pneumonia in the United States. J Infect Dis 2012;205(May (10)):1589–92. Epub 2012 March 23. [3] Scott JA. The preventable burden of pneumococcal disease in the developing world. Vaccine 2007;25(March (13)):2398–405. [4] Hausdorff WP, Bryant J, Paradiso PR, Siber GR. Which pneumococcal serogroups cause the most invasive disease: implications for conjugate vaccine formulation and use, part I. Clin Infect Dis 2000;30(January (1)):100–21. [5] Black S, Shinefield H, Fireman B, Lewis E, Ray P, Hansen JR, et al. Efficacy, safety and immunogenicity of heptavalent pneumococcal conjugate vaccine in children. Northern California Kaiser Permanente Vaccine Study Center Group. Pediatr Infect Dis J 2000;19(3):187–95. [6] O’Brien KL, Moulton LH, Reid R, Weatherholtz R, Oski J, Brown L, et al. Efficacy and safety of seven-valent conjugate pneumococcal vaccine in American Indian children: group randomised trial. Lancet British Edition 2003;362(9381):355–61. [7] O’Brien KL, Dagan R, Makela P. Nasopharyngeal carriage. In: Siber G, Klugman P, Makela P, editors. Pneumococcal vaccines: the impact of conjugate vaccines. Washington, DC: ASM Press; 2008. [8] Rosen JB, Thomas AR, Lexau CA, Reingold A, Hadler JL, Harrison LH, et al. Geographic variation in invasive pneumococcal disease following pneumococcal conjugate vaccine introduction in the United States. Clin Infect Dis 2011;53(July (2)):137–43. [9] Direct and indirect effects of routine vaccination of children with 7-valent pneumococcal conjugate vaccine on incidence of invasive pneumococcal disease—Unites States, 1998–2003. MMWR 2005;54(36):893–7. [10] Simonsen L, Taylor RJ, Young-Xu Y, Haber M, May L, Klugman KP. Impact of pneumococcal conjugate vaccination of infants on pneumonia and influenza hospitalization and mortality in all age-groups in the United States. MBio 2011;2(1):25. [11] Global Vaccine Introduction Status. Vaccine Information Management System (VIMS). Accelerated Vaccine Initiative March, 2013 report. Johns Hopkins Bloomberg School of Public Health. http://www.jhsph.edu/research/ centers-and-institutes/ivac/vims/IVAC-VIMS-Report-2013-03.pdf [12] Recommendations to assure the quality, safety and efficacy of pneumococcal conjugate vaccines. In: WHO Expert Committee on Biological Standardization, 60th meeting of the WHO Expert Committee on Biological Standardization. 2009. [13] O’Brien KL, Millar E, Zell E, Bronsdon M, Weatherholtz R, Reid R, et al. Effect of pneumococcal conjugate vaccine on nasopharyngeal colonization among immunized and unimmunized children. J Infect Dis 2007;196: 1211–20. [14] Dagan R, Givon-Lavi N, Fraser D, Lipsitch M, Siber GR, Kohberger R. Serum serotype-specific pneumococcal anticapsular immunoglobulin G concerntrations after immunization with a 9-valent conjugate pneumococcal vaccine correlate with nasopharyngeal acquisition of pneumococcus. J Infect Dis 2005;192(3):367–76. [15] Fraser D, Dagan R, Givon-Lavi N, Sikuler-Cohen M, Janco J, Chang I, et al. Density of nasopharyngeal colonization (NP-col) of vaccine-type and non-vaccine

144

[16]

[17]

[18]

[19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27]

[28] [29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

S.M. Davis et al. / Vaccine 32 (2014) 133–145 type pneumococci after administration of 9-valent CRM pneumococcal conjugate vaccine, abstr. P97. In: Second international symposium on pneumococci pneumococcal diseases. 2000. Simell B, Auranen K, Kayhty H, Goldblatt D, Dagan R, O’Brien KL. Pneumococcal Carriage Group (PneumoCarr). The fundamental link between pneumococcal carriage and disease. Expert Rev Vaccines 2012;11(7):841–55. Isaacman D, Fletcher MA, Fritzell B, Ciuryla V, Schranz J. Indirect effects associated with widespread vaccination of infants with heptavalent pneumococcal conjugate vaccine. Vaccine 2007;25(13):2420–7. Loo JD, Conklin L, Deloria Knoll M, Fleming-Dutra KE, Park DE, Kirk J, et al. Methods for a systematic review of pneumococcal conjugate vaccine dosing schedules. Pediatr Infect Dis J 2013. Submitted. WHO vaccine preventable diseases monitoring system: immunization schedules by antigen, Selection Centre [accessed 05.02.11] at http://apps.who.int/ immunization monitoring/en/globalsummary/ScheduleResult.cfm Vaccine Information Management System [accessed March, 2011] at https:// apps4.jhsph.edu/Vaccine/Login.aspx?ReturnUrl=%2fVaccine%2fDefault.aspx Deloria-Knoll M, Park D, Johnson TS, Chandir S, Nonyane BA, Conklin L. Effect of pneumococcal conjugate vaccine dosing schedule on immunogenicity: a systematic review. Pediatr Infect Dis J 2013. Submitted. Conklin L, Loo J, Kirk J, Fleming-Dutra K, Deloria Knoll M, Park D, et al. Systematic review of the effects of pneumococcal vaccine dosing schedules on vaccine-type invasive pneumococcal disease among young children worldwide, 1994–2011. Pediatr Infect Dis J 2013. Submitted. Loo J, Conklin L, Fleming-Dutra K, Deloria-Knoll M, Park M, Kirk J, et al. Systematic review of pneumococcal conjugate vaccine dosing schedules on pneumonia. Pediatr Infect Dis J 2013. Submitted. Fleming-Dutra K, Conklin L, Loo J, Deloria-Knoll M, Park D, Kirk J, et al. Differential effect of pneumococcal conjugate vaccine dosing schedules on vaccine-type nasopharyngeal carriage: a systematic review. Pediatr Infect Dis J 2013. Submitted. Walker N, Fischer-Walker C, Bryce J, Bahl R, Cousens S. CHERG Review Groups on Intervention Effects. Standards for CHERG reviews of intervention effects on child survival. Int J Epidemiol 2010;39(April (Suppl 1)):i21–31. Shafinoori S, Ginocchio CC, Greenberg AJ, Yeoman E, Cheddie M, Rubin LG. Impact of pneumococcal conjugate vaccine and the severity of winter influenza-like illnesses on invasive pneumococcal infections in children and adults. Pediatr Infect Dis J 2005;24(1):10–6. Tsai CJ, Griffin MR, Nuorti JP, Grijalva CG. Changing epidemiology of pneumococcal meningitis after the introduction of pneumococcal conjugate vaccine in the United States. Clin Infect Dis 2008;46(11):1664–72. Shah SS, Ratner AJ. Trends in invasive pneumococcal disease-associated hospitalizations. Clin Infect Dis 2006;42(January (1)):e1–5. Lin SH, Tan CK, Lai CC, Wang CY, Hsueh PR. Declining incidence of nonbacteremic pneumococcal pneumonia in hospitalized elderly patients at a tertiary care hospital after the introduction of pneumococcal vaccines in Taiwan, 2004–2008. J Am Geriatr Soc 2010;58(1):195–6. Flasche S, Robertson C, Diggle M, Lamb K, McMenamin J, Choi Y. Trends in serotypes among cases of invasive pneumococcal disease (IPD) in Scotland after introduction of PCV7. In: Abstract presented at international symposium on pneumococci and pneumococcal diseases. 2010, 2010 March 14–18, Tel Aviv, Israel. Givon LN, Fraser D, Dagan R. Vaccination of day-care center attendees reduces carriage of Streptococcus pneumoniae among their younger siblings. Pediatr Infect Dis J 2003;22(6):524–31. Cheung YB, Zaman SM, Nsekpong ED, Van Beneden CA, Adegbola RA, Greenwood B, et al. Nasopharyngeal carriage of Streptococcus pneumoniae in Gambian children who participated in a 9-valent pneumococcal conjugate vaccine trial and in their younger siblings. Pediatr Infect Dis J 2009;28(11): 990–5. Sa-Leao R, Nunes S, Brito-Avo A, Frazao N, Simoes AS, Crisostomo MI, et al. Changes in pneumococcal serotypes and antibiotypes carried by vaccinated and unvaccinated day-care centre attendees in Portugal, a country with widespread use of the seven-valent pneumococcal conjugate vaccine. Clin Microbiol Infect 2009;15(11):1002–7. Dias R, Canica M. Invasive pneumococcal disease in Portugal prior to and after the introduction of pneumococcal heptavalent conjugate vaccine. FEMS Immunol Med Microbiol 2007;51:35–42. Huang S, Platt R, Rifas-Shiman S, Pelton S, Goldmann D, Finkelstein J. Post-PCV7 changes in colonizing pneumococcal serotypes in 16 Massachusetts communities, 2001 and 2004. Pediatrics 2005;116(3):e408–13, http://dx.doi.org/10.1542/peds.2004-2338. Weatherholtz R, Millar EV, Moulton LH, Reid R, Rudolph K, Santosham M, et al. Invasive pneumococcal disease a decade after pneumococcal conjugate vaccine use in an American Indian population at high risk for disease. Clin Infect Dis 2010;50(May (9)):1238–46. Foster D, Walker AS, Griffiths D, Peto T, Haworth E, Crook D. Invasive pneumococcal disease—initial impact of the conjugate vaccine in the Oxfordshire region of the UK. In: Abstract 3-122 presented at the 6th international symposium on pneumococci and pneumococcal diseases. 2008. Kellner JD, Church DL, MacDonald J, Scheifele D, Tyrrell GJ, Vanderkooi OG. Serotype replacement in invasive Streptococcus pneumoniae infections after introduction of 7-valent pneumococcal conjugate vaccine (PCV7) in Canada: Implications for an expanded 13-valent conjugate vaccine. In: Abstract PO10.26 presented at 5th international symposium on pneumococci and pneumococcal disease. 2006.

[39] Van der Linden M, Reinert RR, Imohl M. Herd immunity among adults as a consequence of the national immunization programme with pneumococcal conjugate vaccine for children in Germany. In: Abstract presented at the 7th international symposium on pneumococci and pneumococcal diseases. 2010. [40] Munoz-Almagro C, Jordan I, Gene A, Latorre C, Garcia-Garcia J, Pallares R. Emergence of invasive pneumococcal disease caused by nonvaccine serotypes in the era of 7-valent conjugate vaccine. Clin Infect Dis 2008;46:174–82. [41] Grau I, Ardanuy C, Linares J, Podzamczer D, Schulze MH, Pallares R. Trends in mortality and antibiotic resistance among HIV-infected patients with invasive pneumococcal disease. HIV Med 2009;10(8):488–95. [42] Perez-Trallero E, Marimon JM, Ercibengoa M, Vicente D, Perez-Yarza EG. Invasive Streptococcus pneumoniae infections in children and older adults in the north of Spain before and after the introduction of theheptavalent pneumococcal conjugate vaccine. Eur J Clin Microbiol Infect Dis 2009;28(7):731–8. [43] Aguiar SI, Serrano I, Pinto FR, Melo-Cristino J, Ramirez M. Changes in Streptococcus pneumonia serotypes causing invasive disease with non-universal vaccination coverage of the seven-valent conjugate vaccine. Clin Microbiol Infect 2008;14(9):835–43. [44] Fenoll A, Aguilar L, Granizo JJ, Gimenez MJ, Aragoneses-Fenoll L, Mendez C, et al. Has the licensing of respiratory quinolones for adults and the 7-valent pneumococcal conjugate vaccine (PCV-7) for children had herd effects with respect to antimicrobial non-susceptibility in invasive Streptococcus pneumoniae? J Antimicrob Chemother 2008;62(6):1430–3. [45] Lambertsen L, Valentiner-Branth P, Harboe ZB, St-Martin G, Gilsmann S, Andersen PHS, et al. Changes in the serotype distribution of invasive pneumococci after introduction of the PCV7 in Denmark October 2007. In: Abstract presented at 7th international symposium on pneumococci and pneumococcal diseases. 2010. [46] Scott JR, Millar EV, Lipsitch M, Moulton L, Weatherholtz R, Perilla M. Impact of more than a decade of pneumococcal conjugate vaccine use on carriage and invasive potential in Native American communities. J Infect Dis 2012;205(2):280–8. [47] Ho PL, Chiu SS, Chan MY, Ang I, Chow KH, Lau YL. Changes in nasopharyngeal carriage and serotype distribution of antibiotic-resistant Streptococcus pneumonia before and after the introduction of 7-valent pneumococcal conjugate vaccine in Hong Kong. Diag Microb Infect Dis 2011;71:327–34. [48] Spijkerman J, van Gils EJM, Veenhoven RH, Hak E, Yzerman EPF, van der Ende A, et al. Carriage of Streptococcus pneumoniae 3 years after start of vaccination program, the Netherlands. Emerg Infect Dis 2011 [serial on the Internet] http://dx.doi.org/10.3201/eid1704101115 [49] Burgos J, Penaranda M, Payeras A, Villoslada A, Curran A, Garau M, et al. Invasive pneumococcal disease in HIV-infected adults: clinical changes after the introduction of the pneumococcal conjugate vaccine in children. J Acquir Immune Defic Syndr 2012;59(1):31–8. [50] Hanna JH, Humphreys JL, Murpy DM, Smith HV. Invasive pneumococcal disease in non-indigenous people in north Queensland, 2001–2009. Med J Australia 2010;193:392–6. [51] Johnson DR, D’Onise KD, Holland RA, Raupach JCA, Koehler AP. Pneumococcal disease in South Australia: vaccine success but no time for complacency. Vaccine 2012;30:2206–11. [52] Thigpen MC, Whitney CG, Messonier NE, Zell ER, Lynfield R, Hadler JL, et al. Bacterial Meningitis in the United States, 1998–2007. N Engl J Med 2011;364:2016–25. [53] Rosen JB, Thomas AR, Lexau CA, Reingold A, Hadler J, Harrison LH, et al. Geographic variation in invasive pneumococcal disease following pneumococcal conjugate vaccine introduction in the United States. Clin Infect Dis 2011;53(2):137–43. [54] Sahni V, Naus M, Hoang L, Tyrrell GJ, Martin I, Patrick DM. The epidemiology of invasive pneumococcal disease in British Columbia following implementation of an infant immunization program: increases in herd immunity and replacement disease. Can J Public Health 2012;103(1):29–33. [55] Leal JL, Vanderkooi OG, Church DL, MacDonald J, Tyrrell GJ, Kellner JD. Eradication of invasive pneumococcal disease due to the seven-valent pneumococcal conjugate vaccine serotypes in Calgary, Alberta. Pediatr Infect Dis J 2012;31:e169–75. [56] Miller E, Andrews NJ, Waight PA, Slack MPE, George RC. Herd immunity and serotype replacement 4 years after seven-valent pneumococcal conjugate vaccination in England and Wales: an observational cohort study. Lancet Infect Dis 2011;11:760–8. [57] Van der Linden M, Wei S, Falkenhorst G, Siedler A, Imohl M, von Kries R. Four years of universal pneumococcal conjugate infant vaccination in Germany: impact on incidence of invasive pneumococcal disease and serotype distribution in children. Vaccine 2012;30:5880–5. [58] Ingels H, Rasmussen J, Andersen PH, Harboe ZB, Glismann S, Konradsen H, et al. Impact of pneumococcal vaccination in Denmark during the first 3 years after PCV introduction in the childhood immunization programme. Vaccine 2012;30:3944–50. [59] Pirez MC, Algorta GA, Cedres A, Sobrero H, Varela A, Giachetto G, et al. Impact of universal pneumococcal vaccination on hospitalizations for pneumonia and meningitis in children in Montevideo, Uruguay. Pediatr Infect Dis J 2011;30:669–74. [60] Jacups SP, Cheng A. The epidemiology of community acquired bacteremic pneumonia, due to Streptococcus pneumonia, in the top end of the Northern Territory, Australia – over 22 years. Vaccine 2011;29:5386–92. [61] Hennessy TW, Singleton RJ, Bulkow LR, Bruden DL, Hurlburt DA, Parks D, et al. Impact of heptavalent pneumococcal conjugate vaccine on invasive disease,

S.M. Davis et al. / Vaccine 32 (2014) 133–145

[62]

[63]

[64]

[65]

[66]

[67]

[68]

[69]

[70]

antimicrobial resistance and colonization in Alaska Natives: progress towards elimination of a health disparity. Vaccine 2005;23(48–49):5464–73. Singleton R, Hennessy TW, Bulkow LR, Hammitt LL, Zulz T, Hurlburt DA. Invasive pneumococcal disease caused by nonvaccine serotypes among Alaska Native children with high levels of 7-valent pneumococcal vaccine coverage. JAMA 2007;297(16):1784–92. Hammitt LL, Bruden DL, Butler JC, Baggett HC, Hurlburt DA, Reasonover A, et al. Indirect effect of conjugate vaccine on adult carriage of Streptococcus pneumoniae: an explanation of trends in invasive pneumococcal disease. J Infect Dis 2006;193(11):1487–94. Hanna JN, Humphreys JL, Murphy DM. Invasive pneumococcal disease in Indigenous people in north Queensland, 1999–2004. Med J Australia 2006;184(3):118–21. Hanna JN, Humphreys JL, Murphy DM. Invasive pneumococcal disease in Indigenous people in north Queensland: an update, 2005–2007. Med J Australia 2008;189(1):43–6. O’Brien KL, Reid RR, Moulton L, Katz S, Weatherholtz R, Donaldson C, et al. Is use of conjugate pneumococcal vaccine changing invasive pneumococcal disease among Navajo and White Mountain Apache populations? Abstract presented at 3rd international symposium on pneumococci and pneumococcal disease. 2002. O’Brien KL, Weatherholtz R, Millar EV, Watt JP, Reid R, Parkinson A, et al. Replacement invasive pneumococcal disease 9 years after introduction of PCV use among a population at high risk for IPD: the Navajo experience. In: Abstract PO4.17 presented at 5th international symposium on pneumococci and pneumococcal disease. 2006. Craig MJ, Larzalere-Hinton F, Frizell F, Lacapa R, Parkinson A, Leonard L, et al. Preventing pneumococcal disease in the White Mountain Apache American Indian community. In: Abstract SY2.05 presented at 5th international symposium on pneumococci and pneumococcal disease. 2006. Gil PR, San RMJ, Gomez AC, Alvaro MLA, Rivero A, Gil DMA. Epidemiology of pneumococcal meningitis hospitalizations in pediatric population in Spain (1998–2006). Vaccine 2009;27(20):2669–73. Tyrrell GJ, Lovgren M, Chui N, Minion J, Garg S, Kellner JD, et al. Serotypes and antimicrobial susceptibilities of invasive Streptococcus pneumoniae pre- and

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78] [79]

145

post-seven valent pneumococcal conjugate vaccine introduction in Alberta, Canada, 2000–2006. Vaccine 2009;27(27):3553–60. Whitney CG, Farley MM, Hadler J, Harrison LH, Lexau C, Reingold A, et al. Decline in invasive pneumococcal disease in the U.S. in 2000: An effect of pneumococcal conjugate vaccine? Abstracts of the interscience conference on antimicrobial agents & chemotherapy, vol. 48. 2001. p. 386. Kaplan SL, Mason EOJR, Wald ER, Schutze GE, Bradley JS, Tan TQ, et al. Decrease of invasive pneumococcal infections in children among 8 children’s hospitals in the United States after the introduction of the 7-valent pneumococcal conjugate vaccine. Pediatrics 2004;113(3 Pt 1):443–9. McIntyre PB, Gilmour R, Brown M, Watson M, Bartlett M, Gilbert GL. Invasive pneumococcal disease and the impact of the 7-valent pneumococcal conjugate vaccine in the greater Sydney region NSW 1998–2005. In: Abstract PO4.04 presented at the 5th international symposium on pneumococci and pneumococcal disease. 2006. Lepoutre A, Varon E, Georges S, Gutmann L, Levy-Bruhl D. Impact of pneumococcal vaccination on invasive pneumococcal diseases in France, 2001–2006. Eur Surveill 2008;13(August (35)):1–5. Dubos F, Marechal I, Husson MO, Courouble C, Aurel M, Martinot A. Decline in pneumococcal meningitis after the introduction of the heptavalentpneumococcal conjugate vaccine in northern France. Arch Dis Child 2007;92:1009–12. Feikin D, Moore M, Serotype Replacement Study Team. A multi-site analysis of serotype-specific changes in invasive pneumococcal disease after conjugate vaccine introduction among children <5 years old. In: 8th international symposium on pneumococci and pneumococcal disease. 2012. Changing epidemiology of pneumococcal serotypes after introduction of conjugate vaccine: July 2010 report. Wkly Epidemiol Rec 2010;43: 434–6. Weinberger DM, Malley R, Lipsitch M. Serotype replacement in disease after pneumococcal vaccination. Lancet 2011;378(9807):1962–73. Grijalva CG, Nuorti JP, Arbogast PG, Martin SW, Edwards KM, Griffin MR. Decline in pneumonia admissions after routine childhood immunisation with pneumococcal conjugate vaccine in the USA: a time-series analysis. Lancet 2007;369(9568):1179–86.