Characterization of Streptococcus pneumoniae detected in clinical respiratory tract samples in southern Sweden 2 to 4 years after introduction of PCV13

Open AccessPublished:May 29, 2021DOI:https://doi.org/10.1016/j.jinf.2021.05.031

      Highlights

      • Serotype 19A but not 3 declines in mucosal infections after introduction of PCV13.
      • Serotype 19A remains present and may reemerge in pediatric IPD if PCV13 is halted.
      • Respiratory and invasive infections are caused by different nonvaccine serotypes.
      • Certain replacement serotypes are associated with important antimicrobial resistance.

      Summary

      Objective

      To determine the serotype distribution and antimicrobial resistance of Streptococcus pneumoniae associated with mucosal infections in patients of all ages, 2 to 4 years after the transition from a 10-valent pneumococcal conjugate vaccine (PCV10) to PCV13 in the childhood immunization programme.

      Methods

      Background information and antimicrobial susceptibility data regarding all respiratory tract, middle ear, and conjunctival samples positive for growth of S. pneumoniae (n = 2,131) were collected during 18 months in 2016–2018. Available corresponding bacterial isolates were serotyped by PCR and/or antisera (n = 1,858).

      Results

      In total, 17% of isolates were covered by PCV13, predominantly represented by serotypes 3 (9%) and 19A (5%). The most common nonvaccine serotypes were 11A (10%), 23B (10%), 15A (6%) and 35F (5%). Isolates exhibiting serotype 15A or 23B were often multidrug-resistant (21%) or penicillin nonsusceptible (38%), respectively.

      Conclusions

      The overall proportion of serotype 19A was halved compared to a previous observation period when PCV10 was used (years 2011–2013), suggesting herd protection related to PCV13. The proportion of serotype 3 was, however, unchanged. Despite most nonvaccine serotypes causing mucosal infections have a low invasive potential, certain antibiotic resistant serotypes may pose a clinical problem.

      Keywords

      Introduction

      The respiratory tract commensal Streptococcus pneumoniae is a transient colonizer of the human nasopharynx and an important etiology of acute otitis media (AOM), community acquired pneumonia (CAP) and invasive pneumococcal disease (IPD), predominantly causing infections in children <5 years of age and the elderly [
      GBD 2016 Lower Respiratory Infections Collaborators
      Estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory infections in 195 countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study.
      ]. In Sweden, the 7-valent pneumococcal conjugate vaccine (PCV7) was gradually introduced in the national childhood immunization programme between 2007 and 2009. In 2010, PCV10 and PCV13 replaced that vaccine, with different counties using one or the other. Decreased incidences of IPD, CAP and AOM were seen in children after vaccine introduction, while herd protection against IPD and CAP among the elderly was limited [
      • Naucler P.
      • Galanis I.
      • Morfeldt E.
      • Darenberg J.
      • Ortqvist A.
      • Henriques-Normark B
      Comparison of the impact of pneumococcal conjugate vaccine 10 or pneumococcal conjugate vaccine 13 on invasive pneumococcal disease in equivalent populations.
      ,
      • Galanis I.
      • Lindstrand A.
      • Darenberg J.
      • Browall S.
      • Nannapaneni P.
      • Sjostrom K.
      • et al.
      Effects of PCV7 and PCV13 on invasive pneumococcal disease and carriage in Stockholm, Sweden.
      ,
      • Johansson Kostenniemi U.
      • Norman D.
      • Sellin M.
      • Silfverdal S.A
      Sustained reductions of invasive infectious disease following general infant Haemophilus influenzae type b and pneumococcal vaccination in a Swedish Arctic region.
      ,
      • Naucler P.
      • Henriques-Normark B.
      • Hedlund J.
      • Galanis I.
      • Granath F.
      • Ortqvist A.
      The changing epidemiology of community-acquired pneumonia: nationwide register-based study in Sweden.
      ,
      • Gisselsson-Solen M.
      Trends in otitis media incidence after conjugate pneumococcal vaccination: a national observational study.
      ,
      • Hanquet G.
      • Krizova P.
      • Valentiner-Branth P.
      • Ladhani S.N.
      • Nuorti J.P.
      • Lepoutre A.
      • et al.
      Effect of childhood pneumococcal conjugate vaccination on invasive disease in older adults of 10 European countries: implications for adult vaccination.
      ]. A similar situation was observed in several other European countries since serotype replacement occurs in this age group, leading to a reduced effect on the overall disease incidence [
      • Hanquet G.
      • Krizova P.
      • Valentiner-Branth P.
      • Ladhani S.N.
      • Nuorti J.P.
      • Lepoutre A.
      • et al.
      Effect of childhood pneumococcal conjugate vaccination on invasive disease in older adults of 10 European countries: implications for adult vaccination.
      ].
      We have previously described the serotype distribution of pneumococci isolated from clinical, disease-related, upper respiratory tract (URT) specimens prior to and after introduction of PCV7 (2009) and PCV10 (2010) in Skåne county [
      • Littorin N.
      • Ahl J.
      • Uddén F.
      • Resman F.
      • Riesbeck K.
      Reduction of Streptococcus pneumoniae in upper respiratory tract cultures and a decreased incidence of related acute otitis media following introduction of childhood pneumococcal conjugate vaccines in a Swedish county.
      ]. A clear decline of PCV10 serotypes and an overall decrease of the number of samples positive for S. pneumoniae was observed in 2011–2013 compared with 2007–2008. In contrast, during the same time period, the proportions of PCV13 serotype 19A, a serotype associated with severe infections and antibiotic resistance [
      • Isturiz R.
      • Sings H.L.
      • Hilton B.
      • Arguedas A.
      • Reinert R.R.
      • Jodar L.
      Streptococcus pneumoniae serotype 19A: worldwide epidemiology.
      ], and several non-vaccine types (NVTs) increased.
      The aim of this study was to describe the spectrum of disease-related S. pneumoniae present in the population of Skåne county, southern Sweden, after the transition from using PCV10 to PCV13 in the childhood immunization programme (May 2014), by characterizing pneumococcal isolates detected in clinical respiratory tract cultures (including middle ear and conjunctival samples). The serotype distribution and antimicrobial resistance rates of isolated pneumococci were assessed. Additionally, the serotype distribution was compared to that of pneumococcal isolates causing IPD within the county during the same years. While pneumococci causing IPD are monitored nationally in Sweden, the body of knowledge is limited regarding vaccine effects on mucosal infections, and what serotypes that cause them and their corresponding antimicrobial resistance patterns. In addition to PCV13 serotypes 3 and 19A, which were both prevalent previously [
      • Littorin N.
      • Ahl J.
      • Uddén F.
      • Resman F.
      • Riesbeck K.
      Reduction of Streptococcus pneumoniae in upper respiratory tract cultures and a decreased incidence of related acute otitis media following introduction of childhood pneumococcal conjugate vaccines in a Swedish county.
      ], emerging NVTs were objectives of special interest in our present study.

      Methods

       Study setting

      This study was conducted in southern Sweden (Skåne county) between 1st October 2016 and 31st March 2018. The total population of the county increased from approximately 1,320,000 to 1,349,000 inhabitants during the study period (https://www.scb.se/, accessed 01–03–2020). PCV7 was used in the childhood immunization program in Skåne county between January 2009 and May 2010. Thereafter, PCV10 was used until 1st May 2014 when it was replaced by PCV13. Since 2010, the vaccine uptake of PCV among eligible children nationally as well as in Skåne county has been approximately 97–98% (https://www.folkhalsomyndigheten.se, accessed 30–03–2021). Indications for microbiological sampling of the respiratory tract, middle ear discharge or conjunctiva in primary care settings in Sweden are complicated and/ or therapy-resistant AOM, sinusitis or conjunctivitis, in addition to suspected or confirmed CAP, although any culture may be collected at the discretion of the referring physician. At emergency departments and hospital wards, nasopharyngeal and lower respiratory tract (LRT) culture samples are primarily obtained from patients with suspected LRT infection.

       Data collection

      Data on all respiratory tract, middle ear and conjunctival culture samples positive for S. pneumoniae were obtained from Clinical microbiology (Laboratory medicine, Lund) (n = 2,288). Information included age (range 0–100 years) and sex (51% females) of the sampled patient, collection date, referring unit (69% outpatient settings) and sample type. Specimens from the nasopharynx (n = 1,384), middle ear discharge (n = 85), conjunctival secretions (n = 72), and sinus/nasal discharge (n = 44) were considered URT samples. Samples from the LRT comprised sputum (n = 547) and samples from trachea/bronchi (n = 150). Five samples were of undefined respiratory tract secretions. If multiple samples positive for S. pneumoniae were collected within two months from the same patient, and the serotype of the corresponding isolates was identical, only one sample was further analyzed (n = 151 samples excluded). Serotype data regarding pneumococcal isolates from patients with IPD in Skåne county during 2016 to 2018 were obtained from the Public Health Agency of Sweden (n = 519).

       Collection of bacterial isolates, culture conditions and species identification

      Pneumococcal isolates (n = 1,858), corresponding to 87% of the included culture samples, were stored at −80 °C in glycerol-supplemented horse serum after identification by standard methods. For subsequent analyses, isolates were cultured on blood agar plates for 18 h at 37 °C in 5% CO2. Five non-pneumococcal isolates, which had initially been classified as S. pneumoniae, were excluded from the study.

       Antimicrobial susceptibility testing (AST)

      Pneumococcal isolates were screened for nonsusceptibility to oxacillin, erythromycin, clindamycin, tetracycline and trimethoprim/sulfamethoxazole using disk diffusion tests (Thermo Fisher Scientific, Waltham, MA). Minimum inhibitory concentrations (MICs) of all oxacillin-resistant isolates were determined with gradient tests for benzylpenicillin, ampicillin and cefotaxime (Etest; BioMérieux, Marcy-l'Étoile, FR). Additional tests for other antibiotics were performed based on the clinical situation. All results from AST were interpreted according to current EUCAST breakpoints at the time of testing (https://eucast.org/clinical_breakpoints/; Clinical breakpoints - bacteria v6.0–8.0). Isolates were considered putatively multidrug-resistant (MDR) or extensively drug-resistant (XDR) if nonsusceptible to ≥3 or ≥5 antimicrobials during screening, respectively. MICs of XDR isolates were determined with broth microdilution (Sensititre Streptococcus STP6F AST Plate; Thermo Fisher Scientific).

       Serotyping of pneumococci

      Serotyping was performed using a multiplex polymerase chain reaction (PCR) comprising 6 sequential reaction in combination with latex agglutination and the Quellung reaction as described [
      • Uddén F.
      • Filipe M.
      • Reimer A.
      • Paul M.
      • Matuschek E.
      • Thegerstrom J.
      • et al.
      Aerobic bacteria associated with chronic suppurative otitis media in Angola.
      ]. Isolates that were negative twice in PCRs were serotyped with the ImmuLex Pneumotest Kit (SSI Diagnostica, Copenhagen, DK) and Neufeldt antisera. Isolates that were negative for cpsA in PCR twice, or could not be determined through latex agglutination and the Quellung reaction, were considered nontypeable (NT).

       Statistical analyses

      All statistical tests were performed in SPSS Statistics 26 (IBM, Armonk, NY). To compare proportion between two groups, the Chi-square test and, if the predicted value in any cell was <5, Fisher's exact test were used. To investigate the association of clinical and demographical variables with the isolation of PCV13 serotypes, univariate and multivariate logistic regression analyses were performed. In these analyses, patient age was converted to a dichotomous variable (“PCV13 eligible age”) based on birthdate prior to or after the introduction on PCV13 (1st May 2014). Serotype-specific IPD odds ratios (ORs) were defined as (a1b1)/(a2b2), where a1 is the number of IPD isolates exhibiting the serotype, b1 the number of IPD isolates not exhibiting the serotype, a2 the number of respiratory tract isolates exhibiting the serotype, and b2 the number of respiratory tract isolates not exhibiting the serotype. Confidence intervals (CI) for frequencies and ORs were calculated as described by Kirkwood and Sterne [
      • Kirkwood B.R.
      • Sterne J.A.C.
      Essential medical statistics.
      ]. Two-tailed p-values < 0.05 were considered statistically significant.

       Ethical considerations

      The project was approved by the local ethics committee (Regionala etikprövningsnämnden i Lund) upon its initiation (approval no. 2012/286) and was updated (approval no. 2016/752) to include the current study.

      Results

       Clinical samples included

      Patient samples were collected 2 to 4 years after the introduction of PCV13 in the national childhood immunization programme. A total of 2,131 samples positive for growth of S. pneumoniae were included in our analyses. Background characteristics of the samples and corresponding patients are summarized in Table 1. In addition, frequencies of S. pneumoniae-positive URT and LRT samples normalized by population size, and the number of samples referred from different health care settings are available in Supplemental Table 1.
      Table 1Background characteristics of samples positive for S. pneumoniae and corresponding patients.
      Dataa for:
      CharacteristicsSamples positive for S. pneumoniaeSamples with serotyped S. pneumoniae
      No. of samples2,1311,858
      Patient age (years)b43 (4–67)42 (3–67)
      Age group
      <2 (years)350 (16)311 (17)
      2–4212 (10)195 (11)
      5–9107 (5)96 (5)
      10–1950 (2)46 (3)
      20–2984 (4)68 (4)
      30–39188 (9)160 (9)
      40–49167 (8)155 (8)
      50–59191 (9)164 (9)
      60–69340 (16)290 (16)
      70–79284 (13)242 (13)
      ≥80158 (7)131 (7)
      Patient sex
      Female1,095 (51)957 (52)
      Male1,036 (49)901 (49)
      Referring department
      Outpatients’ department1,466 (69)1,281 (69)
      Hospital setting665 (31)577 (31)
      Sample type
      Nasopharyngeal swab1,336 (63)1,190 (64)
      Sinus/nasal secretion40 (2)35 (2)
      Middle ear discharge78 (4)71 (4)
      Conjunctival sample66 (3)61 (3)
      Sputum484 (23)399 (22)
      Tracheal/bronchial sample122 (6)98 (5)
      Unspecified airway secretion5 (0.2)4 (0.2)
      aNumbers are n (%) or median (IQR).
      bRange 0–100 years.

       PCV13 serotypes 3 and 19A are still circulating in the population

      The serotype distribution of pneumococcal isolates available for typing is outlined in Fig. 1 (n = 1,858). The overall proportion of serotypes included in PCV13 was 17% (n = 313), and serotypes 3 and 19A comprised 9 and 5%, respectively. In parallel, the percentage of serotypes included in PCV10, and the future vaccines PCV15 (by MSD) and PCV20 (Pfizer) were 3, 21 and 41%, respectively. In Supplemental Tables 2–4, frequencies of all serotypes by age group, sample type and calendar quarter are presented. A multivariate regression analysis revealed that PCV13 serotypes were associated with patients not targeted by PCV13 immunization (p<0.001) and with patient samples collected in hospital settings (p = 0.006) (Table 2).
      Fig. 1
      Fig. 1Serotype distribution of S. pneumoniae isolates (n = 1,858; 87% of all samples positive for growth of S. pneumoniae) detected in clinical respiratory tract samples in Skåne county 2 to 4 years after PCV13 introduction. (A) Overall serotype distribution of pneumococcal isolates grown from upper and lower respiratory tract samples. (B) Serotype distribution of pneumococci detected in samples from patients of different ages. Serotypes included in PCV13 are indicated by the black bars below the x-axes. Non-vaccine types that represent <1% of all isolates are pooled as “Other”. URT, upper respiratory tract; LRT, lower respiratory tract.
      Table 2Univariate and multivariate logistic regression analyses of factors associated with pneumococcal serotypes included in PCV13. The overall frequency of PCV13 serotypes was 17%. Significant associations are indicated in bold. OR, odds ratio; CI, confidence interval; LRT, lower respiratory tract.
      PCV13-included serotype/totalUnivariate analysisMultivariate analysis
      A total of 1,854 samples were included in the multivariate analysis.
      Characteristicsn/n (%)OR95% CIpOR95% CIp
      Patient age
      Continuous variable.
      1.011.005–1.014<0.001
      PCV13 eligible age
      Patients born after 1st May 2014.
      38/440 (9)0.390.28–0.56<0.0010.460.32–0.67<0.001
      Female sex158/957 (17)0.950.75–1.210.69
      Hospital setting125/577 (22)1.611.25–2.07<0.0011.431.11–1.850.006
      LRT sample111/497 (22)1.641.27–2.13<0.0011.311.00–1.710.051
      a A total of 1,854 samples were included in the multivariate analysis.
      b Continuous variable.
      c Patients born after 1st May 2014.

       Serotypes detected in respiratory tract samples rarely cause IPD

      A total of 519 pneumococci were isolated from patients with IPD in Skåne county 2016–2018 and only 2% (n = 11) and 4% (n = 21) of these originated from children younger than 2 and 15 years of age, respectively. The majority (76%) were from individuals aged 60 years or older. To assess whether the serotypes detected in respiratory tract samples also caused IPD, ORs were calculated comparing their frequencies among IPD and respiratory tract samples from patients of all ages as described in Fig. 2. No significant differences were found between ORs calculated with data from patients aged under 15 years, 15–59 years, or over 59 years for any serotype (Supplemental Figure 1). Importantly, most NVTs that were common in respiratory tract infections exhibited IPD/respiratory tract ORs <1.
      Fig. 2
      Fig. 2Serotypes of pneumococcal isolates detected in respiratory tract samples (n = 1,712), compared with samples from sterile sites (IPD, n = 477) in Skåne county 2 to 4 years after PCV13 introduction. The relative frequency of the individual serotypes, as well as PCV13 serotypes pooled, and their IPD/respiratory tract odds ratio are presented. Error bars represent 95% confidence intervals. Serotypes representing <2% of isolates of both respiratory tract and IPD samples are not shown in the figure. CI, confidence interval; IPD, invasive pneumococcal disease; OR, odds ratio.

       A small number of NVTs are responsible for antibiotic resistance

      Antimicrobial nonsusceptibility rates differed considerably between serotypes, and was highest among several PCV13 serotypes and the NVTs 6C, 15A, 23B, 24 and 35B (Fig. 3 and Supplemental Table 5). The detected penicillin nonsusceptible pneumococci (PNSP) (n = 237; 11%) exhibited a low median penicillin MIC (0.13 µg/ml, IQR 0.13–0.25 µg/ml), and only a few PNSP were nonsusceptible to ampicillin (13%) or cefotaxime (3%). MICs of putative XDR pneumococci determined with broth microdilution were as indicated in Supplemental Table 6, these results showed that gradient tests underestimated the penicillin MIC with 1–2 dilution steps in 16 out of 26 (62%) cases.
      Fig. 3
      Fig. 3Antimicrobial nonsusceptibility rates of pneumococci detected in clinical respiratory tract samples in Skåne county, 2 to 4 years after the introduction of PCV13 (n = 2,131). Numbers shown are the percentage of isolates of each serotype, or pooled PCV13 serotypes that were nonsusceptible to the specified substance or exhibited an MDR/XDR phenotype. Gray fields indicate that the nonsusceptibility rate of the corresponding serotype is higher than the overall proportion detected. Underscored numbers indicate a statistically significant difference of the nonsusceptibility rate of a serotype compared to that of all other serotypes combined. Only a minority of isolates (n = 389) were screened for norfloxacin nonsusceptibility and the presented numbers are the nonsusceptibility rates among those tested. Multidrug-resistance and extensive drug-resistance was defined as nonsusceptibility to ≥3 and ≥5 of the tested substances, respectively. Non-vaccine types that represent <1% of all serotyped isolates are pooled as “Other”. CLI, clindamycin; ERY, erythromycin; MDR, multidrug-resistant; NOR, norfloxacin; NT, nontypeable; PCG, benzylpenicillin; PCV, pneumococcal conjugate vaccine; TET, tetracycline; TMP/SMX, trimethoprim-sulfamethoxazole; XDR, extensively drug-resistant.

      Discussion

      We characterized S. pneumoniae isolated from clinical respiratory tract samples at a single laboratory serving an entire county in southern Sweden 2 to 4 years after PCV13 introduction in the childhood immunization programme. Serotypes 3 and 19A were still prevalent, but less in the immunized age group, while remaining PCV13 serotypes were rarely detected. Most frequently isolated NVTs seldomly caused IPD during the study period, but some specific serotypes were often PNSP (15A, 23B and 35B) or MDR/XDR (15A and serogroup 24).
      Pneumococcal serotype epidemiology in relation to PCV use is often investigated through carriage studies of asymptomatic individuals or by examining isolates causing IPD. Studies of clinical non-invasive isolates are scarcer but may be beneficial as they reflect the spectrum of disease-associated pneumococci circulating in the population. A limitation of this approach is although that we cannot fully determine whether the detected isolates represent the true etiology of respiratory infections. Secondly, we did not have information about the diagnoses of the sampled patients. Lastly, penicillin gradient tests were used which may result in underestimated penicillin MIC values, which was indicated by a selection of isolates analyzed with broth microdilution (Supplemental Table 6) [

      European Committee on Antimicrobial Susceptibility Testing. Warning against the use of gradient tests for benzylpenicillin MIC in Streptococcus pneumoniae [Internet]. 2019 Nov 21 [cited 2020 Nov 30]. Available from: https://eucast.org/ast_of_bacteria/warnings/.

      ].
      Compared to our previous report on the years following PCV10 introduction the number of URT samples positive for S. pneumoniae collected from children <2 years decreased from 1,250 (95% CI 1,161–1,341) to 724 (95% CI 651–803) per 100,000 person years (Supplemental Table 1) [
      • Littorin N.
      • Ahl J.
      • Uddén F.
      • Resman F.
      • Riesbeck K.
      Reduction of Streptococcus pneumoniae in upper respiratory tract cultures and a decreased incidence of related acute otitis media following introduction of childhood pneumococcal conjugate vaccines in a Swedish county.
      ]. An in-depth analysis of nasopharyngeal samples collected from children in Skåne county during 14 years also described this decline and showed that it was associated with increasing PCV-coverage [
      • Littorin N.
      • Runow E.
      • Ahl J.
      • Resman F.
      • Riesbeck K.
      Decreased prevalence of Moraxella catarrhalis in addition to Streptococcus pneumoniae in children with upper respiratory tract infection after introduction of conjugated pneumococcal vaccine - A retrospective cohort study.
      ]. This is likely partially explained by the decreased incidences of AOM-related diagnoses and CAP in Swedish children during the study years, but it may also be the result of changes or geographical differences in clinical practice [
      • Naucler P.
      • Henriques-Normark B.
      • Hedlund J.
      • Galanis I.
      • Granath F.
      • Ortqvist A.
      The changing epidemiology of community-acquired pneumonia: nationwide register-based study in Sweden.
      ,
      • Gisselsson-Solen M.
      Trends in otitis media incidence after conjugate pneumococcal vaccination: a national observational study.
      ]. However, no guidelines regarding collection of respiratory cultures have changed since the previous study (2011–2013) and the reduction of S. pneumoniae in nasopharyngeal samples was shown to be independent of the overall reduction of culture referrals supporting the hypothesis that this is an effect of PCV usage [
      • Littorin N.
      • Runow E.
      • Ahl J.
      • Resman F.
      • Riesbeck K.
      Decreased prevalence of Moraxella catarrhalis in addition to Streptococcus pneumoniae in children with upper respiratory tract infection after introduction of conjugated pneumococcal vaccine - A retrospective cohort study.
      ].
      Compared to 2011–2013, the proportion of S. pneumoniae serotype 19A in respiratory tract samples was halved from 10% to 5%, while no overall reduction of serotype 3 (9%) was observed [
      • Littorin N.
      • Ahl J.
      • Uddén F.
      • Resman F.
      • Riesbeck K.
      Reduction of Streptococcus pneumoniae in upper respiratory tract cultures and a decreased incidence of related acute otitis media following introduction of childhood pneumococcal conjugate vaccines in a Swedish county.
      ]. The proportions of both serotypes were lower among children immunized with PCV13 in the current study. This pattern was also observed 2011–2013 regarding serotype 3 (Riesbeck K. et al., unpublished data), suggesting that it might not be an effect of PCV13 implementation. In contrast, the age association of serotype 19A in respiratory tract samples appears to have been established after the switch from PCV10 to PCV13. Taken together, our findings indicate a direct effect of PCV13 on respiratory infections in the vaccinated age group and emerging herd-protection for serotype 19A, while no effect on the occurrence of serotype 3 was seen. In parallel, after the introduction of PCV13 in 2010 in Stockholm (Sweden), serotype 19A decreased in carriage among children <5 years [
      • Lindstrand A.
      • Galanis I.
      • Darenberg J.
      • Morfeldt E.
      • Naucler P.
      • Blennow M.
      • et al.
      Unaltered pneumococcal carriage prevalence due to expansion of non-vaccine types of low invasive potential 8years after vaccine introduction in Stockholm, Sweden.
      ]. Moreover, serotype 3 slightly increased despite it was uncommon at baseline. Studies of PCV10 and PCV13 herd effects on adult noninvasive pneumonia in Iceland and the United States, respectively, also showed declining numbers of infection by the vaccine targeted serotypes, except serotype 3 in the American study [
      • Quirk S.J.
      • Haraldsson G.
      • Hjálmarsdóttir M.Á.
      • van Tonder A.J.
      • Hrafnkelsson B.
      • Bentley S.D.
      • Haraldsson Á.
      • et al.
      Vaccination of Icelandic children with the 10-valent pneumococcal vaccine leads to a significant herd effect among Adults in Iceland.
      ,
      • Mendes R.E.
      • Hollingsworth R.C.
      • Costello A.
      • Jones R.N.
      • Isturiz R.E.
      • Hewlett Jr, D.
      • Farrell D.J.
      Noninvasive Streptococcus pneumoniae serotypes recovered from hospitalized adult patients in the United States in 2009 to 2012.
      ]. Continued transmission of serotype 3 is likely due to a limited vaccine effect on nasopharyngeal colonization [
      • Dagan R.
      • Patterson S.
      • Juergens C.
      • Greenberg D.
      • Givon-Lavi N.
      • Porat N.
      • et al.
      Comparative immunogenicity and efficacy of 13-valent and 7-valent pneumococcal conjugate vaccines in reducing nasopharyngeal colonization: a randomized double-blind trial.
      ]. Similar incomplete herd protection for serotype 19A and serotype 19F, has been proposed as a possible mechanism explaining why they are not eliminated from the population as readily and completely as many other PCV serotypes in some areas, but this has not been confirmed [
      • van der Linden M.
      • Imohl M.
      • Perniciaro S.
      Limited indirect effects of an infant pneumococcal vaccination program in an aging population.
      ]. Persistent circulation of serotype 19A in the population may cause resurgence of its incidence in IPD among children if vaccination against it is halted, which was observed in Belgium after the transition from PCV13 to PCV10 [
      • Desmet S.
      • Lagrou K.
      • Wyndham-Thomas C.
      • Braeye T.
      • Verhaegen J.
      • Maes P.
      • et al.
      Dynamic changes in paediatric invasive pneumococcal disease after sequential switches of conjugate vaccine in Belgium: a national retrospective observational study.
      ]. Bearing this in mind, the high prevalence of this serotype observed requires continued surveillance to detect potential re-emergence in IPD, as PCV10 replaced PCV13 in several Swedish counties, including Skåne county, in 2019 when it was procured nationally.
      The most frequently isolated NVTs (11A, 23B, 15A, 35F, and the closely related 15B and 15C) are common replacement serotypes detected in carriage in Sweden (Stockholm) [
      • Galanis I.
      • Lindstrand A.
      • Darenberg J.
      • Browall S.
      • Nannapaneni P.
      • Sjostrom K.
      • et al.
      Effects of PCV7 and PCV13 on invasive pneumococcal disease and carriage in Stockholm, Sweden.
      ,
      • Lindstrand A.
      • Galanis I.
      • Darenberg J.
      • Morfeldt E.
      • Naucler P.
      • Blennow M.
      • et al.
      Unaltered pneumococcal carriage prevalence due to expansion of non-vaccine types of low invasive potential 8years after vaccine introduction in Stockholm, Sweden.
      ]. They also increased in adult non-invasive pneumonia after the introduction of PCV10 and PCV13 in Iceland and the USA, respectively [
      • Quirk S.J.
      • Haraldsson G.
      • Hjálmarsdóttir M.Á.
      • van Tonder A.J.
      • Hrafnkelsson B.
      • Bentley S.D.
      • Haraldsson Á.
      • et al.
      Vaccination of Icelandic children with the 10-valent pneumococcal vaccine leads to a significant herd effect among Adults in Iceland.
      ,
      • Mendes R.E.
      • Hollingsworth R.C.
      • Costello A.
      • Jones R.N.
      • Isturiz R.E.
      • Hewlett Jr, D.
      • Farrell D.J.
      Noninvasive Streptococcus pneumoniae serotypes recovered from hospitalized adult patients in the United States in 2009 to 2012.
      ]. While these serotypes have low-medium invasive potential, they may cause a considerable proportion of IPD cases in both children and adults if they become sufficiently prevalent in the population that is evident from the numbers outlined in Fig. 2 [
      • Lindstrand A.
      • Galanis I.
      • Darenberg J.
      • Morfeldt E.
      • Naucler P.
      • Blennow M.
      • et al.
      Unaltered pneumococcal carriage prevalence due to expansion of non-vaccine types of low invasive potential 8years after vaccine introduction in Stockholm, Sweden.
      ,
      • Balsells E.
      • Dagan R.
      • Yildirim I.
      • Gounder P.P.
      • Steens A.
      • Munoz-Almagro C.
      • et al.
      The relative invasive disease potential of Streptococcus pneumoniae among children after PCV introduction: a systematic review and meta-analysis.
      ]. The majority of IPD cases were, however, caused by PCV13 serotypes (mostly 3 and 19A) and NVTs that were seldom detected in respiratory tract samples (serotypes 8 and 9N, and serogroup 22), which is similar to the situation in Germany among patients aged 60 years or older [
      • van der Linden M.
      • Imohl M.
      • Perniciaro S.
      Limited indirect effects of an infant pneumococcal vaccination program in an aging population.
      ]. Serotype 15A was an exception, being the only NVT that was among the most prevalent in both patient groups. Even if the IPD ORs calculated for serotypes in the current study do not represent invasive disease potential by definition, they largely agree with the invasive disease potentials of serotypes detected in Stockholm 2011–2015 [
      • Lindstrand A.
      • Galanis I.
      • Darenberg J.
      • Morfeldt E.
      • Naucler P.
      • Blennow M.
      • et al.
      Unaltered pneumococcal carriage prevalence due to expansion of non-vaccine types of low invasive potential 8years after vaccine introduction in Stockholm, Sweden.
      ]. This further indicated that the spectrum of serotypes causing mucosal infections is more similar to that found in carriage than in IPD. However, because only a small number of IPD samples were from children, the ORs should be interpreted with caution for this particular age group. Finally, the fact that 76% of the IPD samples used in the analyses were collected from individuals aged 60 years or older clearly illustrates how the age distribution of IPD patients has changed in Sweden since PCV introduction, the incidence declined among children <2 years but was unaltered among the elderly [
      • Galanis I.
      • Lindstrand A.
      • Darenberg J.
      • Browall S.
      • Nannapaneni P.
      • Sjostrom K.
      • et al.
      Effects of PCV7 and PCV13 on invasive pneumococcal disease and carriage in Stockholm, Sweden.
      ].
      In parallel to our results, the association of NVTs 6C, 15A, 23B, 24F and 35B to antibiotic resistance has previously been reported [
      • van der Linden M.
      • Perniciaro S.
      • Imohl M.
      Increase of serotypes 15A and 23B in IPD in Germany in the PCV13 vaccination era.
      ,
      • Richter S.S.
      • Diekema D.J.
      • Heilmann K.P.
      • Dohrn C.L.
      • Riahi F.
      • Doern G.V.
      Changes in pneumococcal serotypes and antimicrobial resistance after introduction of the 13-valent conjugate vaccine in the United States.
      ,
      • Neves F.P.G.
      • Cardoso N.T.
      • Snyder R.E.
      • Marlow M.A.
      • Cardoso C.A.A.
      • Teixeira L.M.
      • et al.
      Pneumococcal carriage among children after four years of routine 10-valent pneumococcal conjugate vaccine use in Brazil: the emergence of multidrug resistant serotype 6C.
      ,
      • Adam H.J.
      • Golden A.R.
      • Karlowsky J.A.
      • Baxter M.R.
      • Nichol K.A.
      • Martin I.
      • et al.
      Analysis of multidrug resistance in the predominant Streptococcus pneumoniae serotypes in Canada: the SAVE study, 2011-15.
      ]. Further studies are needed to comment on the clinical significance of the antimicrobial resistance patterns of these serotypes in Sweden. Data from Norway and Germany, however, showed an increase in IPD caused by serotypes 15A and 23B after introduction of PCV13, replacement serotypes that were often MDR or PNSP, respectively [
      • van der Linden M.
      • Perniciaro S.
      • Imohl M.
      Increase of serotypes 15A and 23B in IPD in Germany in the PCV13 vaccination era.
      ,
      • Steens A.
      • Bergsaker M.A.
      • Aaberge I.S.
      • Ronning K.
      • Vestrheim D.F.
      Prompt effect of replacing the 7-valent pneumococcal conjugate vaccine with the 13-valent vaccine on the epidemiology of invasive pneumococcal disease in Norway.
      ]. Serotype 19A has been the subject of concern in many regions since PCV implementation due to antibiotic resistance, which further emphasizes the importance of monitoring this serotype after the replacement of PCV13 with PCV10 in 2019 [
      • van der Linden M.
      • Reinert R.R.
      • Kern W.V.
      • Imohl M.
      Epidemiology of serotype 19A isolates from invasive pneumococcal disease in German children.
      ].
      In conclusion, this study indicated that the disease burden of pneumococcal respiratory tract infections has continued to decline after PCV13 introduction. Furthermore, serotype 19A has become less prevalent among patients of all ages, which suggests a direct as well as herd effect of PCV13, but is still circulating in the population. In contrast, the prevalence of serotype 3 remains unchanged. Importantly, reversion to the use of PCV10 in the childhood immunization program may lead to an increased incidence of pediatric IPD caused by serotype 19A due to the circulating reservoir of this serotype. Finally, the NVTs that dominated in respiratory tract samples are generally low invasive, but certain serotypes, such as 15A and 23B, may constitute a future clinical problem due to high rates of PNSP or multidrug-resistance.

      Declaration of Competing Interest

      Hans-Christian Slotved and Jonas Ahl report grants from Pfizer outside the submitted work. Kristian Riesbeck reports grants from Pfizer, during the conduct of the study; personal fees from GSK, personal fees from MSD, and grants from Pfizer, outside the submitted work. Remaining authors have nothing to disclose.

      Funding

      The study was financially supported by an unrestrained grant from Pfizer, the Anna and Edwin Berger Foundation (KR), the Royal Physiographical Society (FU; Forssman's Foundation), the Skåne County Council's research and development foundation (KR), Swedish Heart Lung Foundation (KR; #20180401 ), and Swedish Research Council (KR; #2019-01053 ).

      Acknowledgments

      We sincerely thank Mrs. Birgitta Andersson for excellent technical assistance with the microbiological analyses. We thank The Public Health Agency of Sweden for providing data on IPD pneumococcal serotypes used in our analyses.

      Access to data

      All data used in the analyses are available from the authors upon request.

      Contribution

      JA and KR conceived and planned the study. FU and KR administered the project. FU, HCS and KF developed methodology. FU performed experiments. FU, ER and HCS analyzed the results. FU drafted the manuscript with support from ER. All authors contributed to reviewing and editing the manuscript, and have approved the final version.

      Appendix. Supplementary materials

      References

        • GBD 2016 Lower Respiratory Infections Collaborators
        Estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory infections in 195 countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study.
        Lancet Infect Dis. 2016; 18 (2018): 1191-1210https://doi.org/10.1016/S1473-3099(18)30310-4
        • Naucler P.
        • Galanis I.
        • Morfeldt E.
        • Darenberg J.
        • Ortqvist A.
        • Henriques-Normark B
        Comparison of the impact of pneumococcal conjugate vaccine 10 or pneumococcal conjugate vaccine 13 on invasive pneumococcal disease in equivalent populations.
        Clin Infect Dis. 2017; 65: 1780-1789https://doi.org/10.1093/cid/cix685
        • Galanis I.
        • Lindstrand A.
        • Darenberg J.
        • Browall S.
        • Nannapaneni P.
        • Sjostrom K.
        • et al.
        Effects of PCV7 and PCV13 on invasive pneumococcal disease and carriage in Stockholm, Sweden.
        Eur Respir J. 2016; 47: 1208-1218https://doi.org/10.1183/13993003.01451-2015
        • Johansson Kostenniemi U.
        • Norman D.
        • Sellin M.
        • Silfverdal S.A
        Sustained reductions of invasive infectious disease following general infant Haemophilus influenzae type b and pneumococcal vaccination in a Swedish Arctic region.
        Acta Paediatr. 2019; 108: 1871-1878https://doi.org/10.1111/apa.14824
        • Naucler P.
        • Henriques-Normark B.
        • Hedlund J.
        • Galanis I.
        • Granath F.
        • Ortqvist A.
        The changing epidemiology of community-acquired pneumonia: nationwide register-based study in Sweden.
        J Intern Med. 2019; 286: 689-701https://doi.org/10.1111/joim.12956
        • Gisselsson-Solen M.
        Trends in otitis media incidence after conjugate pneumococcal vaccination: a national observational study.
        Pediatr Infect Dis J. 2017; 36: 1027-1031https://doi.org/10.1097/INF.0000000000001654
        • Hanquet G.
        • Krizova P.
        • Valentiner-Branth P.
        • Ladhani S.N.
        • Nuorti J.P.
        • Lepoutre A.
        • et al.
        Effect of childhood pneumococcal conjugate vaccination on invasive disease in older adults of 10 European countries: implications for adult vaccination.
        Thorax. 2019; 74: 473-482https://doi.org/10.1136/thoraxjnl-2018-211767
        • Littorin N.
        • Ahl J.
        • Uddén F.
        • Resman F.
        • Riesbeck K.
        Reduction of Streptococcus pneumoniae in upper respiratory tract cultures and a decreased incidence of related acute otitis media following introduction of childhood pneumococcal conjugate vaccines in a Swedish county.
        BMC Infect Dis. 2016; 16: 407https://doi.org/10.1186/s12879-016-1750-5
        • Isturiz R.
        • Sings H.L.
        • Hilton B.
        • Arguedas A.
        • Reinert R.R.
        • Jodar L.
        Streptococcus pneumoniae serotype 19A: worldwide epidemiology.
        Expert Rev Vaccines. 2017; 16: 1007-1027https://doi.org/10.1080/14760584.2017.1362339
        • Uddén F.
        • Filipe M.
        • Reimer A.
        • Paul M.
        • Matuschek E.
        • Thegerstrom J.
        • et al.
        Aerobic bacteria associated with chronic suppurative otitis media in Angola.
        Infect Dis Poverty. 2018; 7: 42https://doi.org/10.1186/s40249-018-0422-7
        • Kirkwood B.R.
        • Sterne J.A.C.
        Essential medical statistics.
        2nd ed. Blackwell Science, Malden, MA2003
      1. European Committee on Antimicrobial Susceptibility Testing. Warning against the use of gradient tests for benzylpenicillin MIC in Streptococcus pneumoniae [Internet]. 2019 Nov 21 [cited 2020 Nov 30]. Available from: https://eucast.org/ast_of_bacteria/warnings/.

        • Littorin N.
        • Runow E.
        • Ahl J.
        • Resman F.
        • Riesbeck K.
        Decreased prevalence of Moraxella catarrhalis in addition to Streptococcus pneumoniae in children with upper respiratory tract infection after introduction of conjugated pneumococcal vaccine - A retrospective cohort study.
        Clin Microbiol Infect. 2020; (S1198-743X(20)30259-7)https://doi.org/10.1016/j.cmi.2020.04.033
        • Lindstrand A.
        • Galanis I.
        • Darenberg J.
        • Morfeldt E.
        • Naucler P.
        • Blennow M.
        • et al.
        Unaltered pneumococcal carriage prevalence due to expansion of non-vaccine types of low invasive potential 8years after vaccine introduction in Stockholm, Sweden.
        Vaccine. 2016; 34: 4565-4571https://doi.org/10.1016/j.vaccine.2016.07.031
        • Quirk S.J.
        • Haraldsson G.
        • Hjálmarsdóttir M.Á.
        • van Tonder A.J.
        • Hrafnkelsson B.
        • Bentley S.D.
        • Haraldsson Á.
        • et al.
        Vaccination of Icelandic children with the 10-valent pneumococcal vaccine leads to a significant herd effect among Adults in Iceland.
        J Clin Microbiol. 2019; 57 (e01766–18)https://doi.org/10.1128/JCM.01766-18
        • Mendes R.E.
        • Hollingsworth R.C.
        • Costello A.
        • Jones R.N.
        • Isturiz R.E.
        • Hewlett Jr, D.
        • Farrell D.J.
        Noninvasive Streptococcus pneumoniae serotypes recovered from hospitalized adult patients in the United States in 2009 to 2012.
        Antimicrob Agents Chemother. 2015; 59: 5595-5601https://doi.org/10.1128/AAC.00182-15
        • Dagan R.
        • Patterson S.
        • Juergens C.
        • Greenberg D.
        • Givon-Lavi N.
        • Porat N.
        • et al.
        Comparative immunogenicity and efficacy of 13-valent and 7-valent pneumococcal conjugate vaccines in reducing nasopharyngeal colonization: a randomized double-blind trial.
        Clin Infect Dis. 2013; 57: 952-962https://doi.org/10.1093/cid/cit428
        • van der Linden M.
        • Imohl M.
        • Perniciaro S.
        Limited indirect effects of an infant pneumococcal vaccination program in an aging population.
        PLoS One. 2019; 14e0220453https://doi.org/10.1371/journal.pone.0220453
        • Desmet S.
        • Lagrou K.
        • Wyndham-Thomas C.
        • Braeye T.
        • Verhaegen J.
        • Maes P.
        • et al.
        Dynamic changes in paediatric invasive pneumococcal disease after sequential switches of conjugate vaccine in Belgium: a national retrospective observational study.
        Lancet Infect Dis. 2020; https://doi.org/10.1016/S1473-3099(20)30173-0
        • Balsells E.
        • Dagan R.
        • Yildirim I.
        • Gounder P.P.
        • Steens A.
        • Munoz-Almagro C.
        • et al.
        The relative invasive disease potential of Streptococcus pneumoniae among children after PCV introduction: a systematic review and meta-analysis.
        J Infect. 2018; 77: 368-378https://doi.org/10.1016/j.jinf.2018.06.004
        • van der Linden M.
        • Perniciaro S.
        • Imohl M.
        Increase of serotypes 15A and 23B in IPD in Germany in the PCV13 vaccination era.
        BMC Infect Dis. 2015; 15: 207https://doi.org/10.1186/s12879-015-0941-9
        • Richter S.S.
        • Diekema D.J.
        • Heilmann K.P.
        • Dohrn C.L.
        • Riahi F.
        • Doern G.V.
        Changes in pneumococcal serotypes and antimicrobial resistance after introduction of the 13-valent conjugate vaccine in the United States.
        Antimicrob Agents Chemother. 2014; 58: 6484-6489https://doi.org/10.1128/AAC.03344-14
        • Neves F.P.G.
        • Cardoso N.T.
        • Snyder R.E.
        • Marlow M.A.
        • Cardoso C.A.A.
        • Teixeira L.M.
        • et al.
        Pneumococcal carriage among children after four years of routine 10-valent pneumococcal conjugate vaccine use in Brazil: the emergence of multidrug resistant serotype 6C.
        Vaccine. 2017; 35: 2794-2800https://doi.org/10.1016/j.vaccine.2017.04.019
        • Adam H.J.
        • Golden A.R.
        • Karlowsky J.A.
        • Baxter M.R.
        • Nichol K.A.
        • Martin I.
        • et al.
        Analysis of multidrug resistance in the predominant Streptococcus pneumoniae serotypes in Canada: the SAVE study, 2011-15.
        J Antimicrob Chemother. 2018; 73: vii12-vvii9https://doi.org/10.1093/jac/dky158
        • Steens A.
        • Bergsaker M.A.
        • Aaberge I.S.
        • Ronning K.
        • Vestrheim D.F.
        Prompt effect of replacing the 7-valent pneumococcal conjugate vaccine with the 13-valent vaccine on the epidemiology of invasive pneumococcal disease in Norway.
        Vaccine. 2013; 31: 6232-6238https://doi.org/10.1016/j.vaccine.2013.10.032
        • van der Linden M.
        • Reinert R.R.
        • Kern W.V.
        • Imohl M.
        Epidemiology of serotype 19A isolates from invasive pneumococcal disease in German children.
        BMC Infect Dis. 2013; 13: 70https://doi.org/10.1186/1471-2334-13-70