Background

We sought to elucidate the perturbation by PCV13 to the Streptococcus pneumoniae strain and serotype composition in Cambodian carriage isolates.

Methods

Pre-PCV13 (01/2013–12/2015, N=258) and the post-PCV13 isolates (01/2016-02/2017, N=432) were sequenced and analysed using PopPUNK(https://github.com/johnlees/PopPUNK) and SeroBA (https://github.com/sanger-pathogens/seroba) to determine strain prevalence and serotype composition.

Results

PCV13 serotypes significantly decreased by Fisher’s exact test (p=0.003[95% Confidence interval 0.45-0.85], OR 0.62) while non-PCV13 serotype significantly increased(p=0.002[1.18-2.26], OR 1.64) in the post-PCV13 populations. There was a significant increase in Simpsons diversity index for both serotype (Welch’s t-test p=0.0059) and strain (p=0.0228) in the post PCV13 population. The isolates were comprised of 44 unique serotypes with 27 pre-PCV and 32 post-PCV13. Significant changes in prevalence were detected in the post-PCV13 populations of serotypes 19F (N=52, 98.1% GPSC1; p=0.02[0.26-0.89], OR 0.48), 23A (N=27, 96.3% GPSC626; p=0.03 [1.04-9.69], OR 2.84), 34 (N=25, 100% GPSC45; p=0.01 [1.35-24], OR 4.55), and 6D (N=8, 87.5% GPSC16; p=0.03[1.19-Inf], OR Inf).

Conclusions

The strain population in Cambodia has been perturbed by the vaccine but had not yet reached equilibrium 24 months following PCV13 introduction. Additional isolate collection is ongoing for detection of trends towards equilibrium post-PCV13 in this population.

Background

Identifying the molecular characteristics of pneumococci associated with disease may inform development of new clinical interventions. We aimed to perform a bacterial genome-wide association study to identify pneumococcal genes associated with carriage among children with pneumonia.

Methods

DNA from nasopharyngeal pneumococcal isolates obtained from Nepalese children admitted to hospital with pneumonia (cases) and healthy community-based children (controls), underwent whole-genome-sequencing on the Wellcome Sanger Institutes core sequencing pipeline. The association of variants from sequences mapped against the S. pneumoniae ATCC700669 genome, with cluster of orthologous groups using a fixed effects model, was performed using a python based sequence element enrichment analysis.

Results

245 case and 597 control isolates were sequenced. 405461 variants were identified and 31708 tested after filtering. 20 variants from colonising bacteria had a strong association (p<10^-8) with pneumonia. 18/20 of these variants were located within the lacE2 gene. The variant with the strongest association, presence of an A allele at position 1066739, was identified in 240/597 (40%) of controls and 150/245 (61%) of cases (p=10^-10).

Conclusions

In this study in Nepal the pneumococcal gene lacE2 was associated with colonisation in children with pneumonia. Studies examining the role of lacE2 in the pathogenesis of pneumococcal pneumonia are needed.

Background

Pneumococcal carriage influences population-wide strain dynamics, but limited data exist on serotype-specific temporal carriage patterns among PCV-vaccinated West African infants.

Methods

Pneumococcus was cultured from nasopharyngeal swabs (n=1, 595) collected from 102 PCV7-exposed infants followed up from birth to 12 months. Serotyping was performed by whole genome sequencing and sweep-latex agglutination. Parametric survival models with constant hazard rates were fitted to estimate carriage dynamics (duration, clearance and acquisition).

Results

The infants were naturally colonised with 60 pneumococcal serotypes with a mean of 7 (range:2-11) serotypes per infant. Carriage dynamics estimates for serotypes 5, 7F, 39, 9A, and 12F are provided here for the first time in infants. There was no correlation between time to first acquisition and carriage duration (ρ=0.06, P=0.709). Serotype prevalence showed a weak correlation with initial acquisition (ρ=0.07, P=0.706), carriage duration (ρ=0.219, P=0.194), and reacquisition times (ρ=0.09, P=0.730). Onset of initial acquisition was longer than the time taken to reacquire serotypes (median: 136.23 vs 26.15 days, P=7.63×10-6). Overall, serotype-specific carriage durations after initial acquisition and reacquisition were significantly different (P=0.020), varying by serotype.

Conclusions

Pneumococcal carriage dynamics among Gambian infants are complex and highly variable by serotype which may have important implications for transmission and invasive disease.

Background

The 10-valent pneumococcal conjugate vaccine (PCV10) was introduced into the Nepalese infant immunisation schedule in August 2015. We aimed to examine how PCV10 introduction in affected pneumococcal lineages.

Methods

DNA from randomly selected nasopharyngeal pneumococcal isolates of healthy community-based Nepalese children in the Kathmandu valley pre- (2009-2014) and post-PCV10 (2017-2018) introduction, underwent whole-genome-sequencing on the Wellcome Sanger Institutes core sequencing pipeline. Isolates were clustered into lineages based on shared sequence and gene content using Population Partitioning Using Nucleotide K-mers (PopPUNK) software.

Results

313 and 284 pre- and post-PCV10 isolates were sequenced. There was a significant reduction in the proportion of PCV10 serotypes when comparing pre 73/313 (23.3%) with post 37/284 (13%) PCV10 samples (p=0.0014). Overall 122 distinct lineages were identified, 98 pre- and 74 post-PCV10. Simpson's index of diversity for the lineages was 0.992 and 0.987 pre- and post-PCV10 respectively. Within the 3 largest PCV10 serotype lineages there were no examples of non-PCV10 serotype isolates pre-vaccination, whereas all 3 lineages contained non-PCV10 serotypes post-vaccination.

Conclusions

PCV10 serotype prevalence significantly declined following PCV10 introduction. However, strain diversity remained high post-PCV10 and there is evidence suggestive of vaccine escape via capsular-switching among lineages possessing predominantly vaccine-covered capsules prior to PCV10 introduction.

Background

Bangladesh has been generating pneumococcal data since last 30 years to make an evidence-based data for vaccine introduction. This study is aimed to make a genomic characterization of pneumococcus isolated from pre-vaccine period.

Methods

Whole-genome sequencing data of total 525 pneumococcus isolated from IPD, during 2002 to 2015, were analyzed using previously established methods.

Results

Overall, 57 serotypes were identified, and most predominant serotypes were 2, 1, 14, 23F, 5, 19F, 12A and 45 which accounted for 50% of isolates. Serotype coverage were 47% for PCV10+6A, 50% for PCV13 and 58% for PCV20. The population was genetically diverse with 108 known and 61 new Sequence Types (STs), encompassing in 89 GPSCs. Among them, GPSC96 (serotype 2, n=66, 11.6%), GPSC 2 (serotype 1, n=48, 9%), GPSC 9 (serotype 14, n=32, 6%) were most predominant. Significant increase in resistance has observed for Erythromycin (0%-60%). Resistance is commonly seen in GPSC 10, 43, 101 and 482 mainly among serotype 19F, 23F, 6B and 7B, respectively.

Conclusions

Pneumococcus in Bangladesh is diverse and different in respect of serotype, ST and GPSC. This data will work as the baseline population to monitor vaccine induced changes in molecular epidemiology.

Background

Factors associated with nasopharyngeal pneumococcal colonization density have not been comprehensively characterized. Age, immunization status, geographical location, population density, season, and/or acute respiratory illness (ARI) may play a role. We assessed longitudinal colonization density patterns in young rural Peruvian children.

Methods

Nasopharyngeal samples were collected monthly from children aged <3 years followed prospectively each week for ARI from May 2009-September 2011. PCV7 was introduced in the region in 2009. Longitudinally-collected samples from a convenience sample of children with >=1 pneumococcus-positive sample underwent density assessment by lytA qPCR. Density values were log₁₀-transformed to reduce skewness. Assessments were stratified by enrollment age.

Results

Pneumococcus was detected in 471/625 (75%) samples from 30 children; 20/30 (67%) were enrolled before age 1. Variability was observed in colonization densities by calendar quarter and enrollment age, with substantial overlap in density levels and trajectories over time among age groups (Figure). Median densities during ARI episodes (n=62) were 5.60 (IQR 4.36-6.24) compared to non-ARI periods (5.00 [3.98-5.93], n=409, p=0.023).

Conclusions

Among these children, density varied by ARI status but did not clearly decrease with age. Future trajectory analyses will assess serotype-specific density colonization patterns and the association of serotype co-colonization, vaccination, and ARI status with pneumococcal density over time.

Background

Serotype 2 was a major cause of pneumococcal pneumonia about 100 years ago and then disappeared. Recently, serotype 2 re-emerged in many countries, including Bangladesh and associated with meningitis. This study aims to understand genomic and epidemiological characteristics of newly emerged serotype 2 strains.

Methods

Whole-genome sequencing was performed on 146 isolates (invasive= 125, carriage= 8 and other= 5, unknown= 8) collected between 1989 and 2017. Data were analyzed for comparative genomics, antimicrobial resistance and molecular typing.

Results

Isolates were from 16 countries, mostly in Asia (n=93), Africa (n=23) and Oceania (n=26). Bangladesh (n=66) and Papua New Guinea (n=26) contributed 63% of the isolates. Among the known clinical conditions, 80% (91/113) were from meningitis. All isolates belonged to GPSC96 lineage and descended from two predominant sequence types: ST74 found in Asia and Africa, and ST1504 found in Papua New Guinea and Israel. Almost all isolates were sensitive to all antibiotics. No significant genetic differences were detected between invasive and carriage isolates.

Conclusions

Our findings don’t explain why the recent increase in serotype 2 occurred but exclude an outbreak or emergence of an antimicrobial-resistant strain as the cause. These isolates have unusually high propensity to be invasive, mostly causing meningitis.

Background

The invasive pneumococcal disease remains one of the leading causes of morbidity and mortality worldwide. In this study, we investigated high-resolution population genetic structure of S. pneumoniae isolates in Moscow, using the genomic definition of pneumococcal lineages (Global Pneumococcal Sequence Clusters (GPSCs)), serotypes and antimicrobial resistance patterns.

Methods

Eighty-seven pneumococcal isolates were recovered from cerebrospinal fluid and nasopharyngeal swabs of patients with meningitis and upper respiratory tract infections, ages one to 93 years in Moscow between 2011-2015. The serotypes and multilocus sequence types (MLSTs) were derived from whole-genome sequencing data using ARIBA and MLSTcheck. GPSCs were assigned by popPUNK. Antibiotic susceptibility was predicted based on genotypes.

Results

Sixty-seven sequence types identified in the collection belonged to 39 clonal complexes (CCs) and 10 singletons. Overall, 42 GPSCs were identified. The two prevalent lineages were GPSC1 (CC320) and GPSC7 (CC437). Twenty-two serotypes were found and their associated GPSCs are shown in Figure 1. The major GPSC contributing to multidrug resistance was GPSC1, which expressed 19F/19A serotypes (Figure 2).

Conclusions

The pneumococcal collection showed high diversity in population structure. Ongoing surveillance is needed to monitor the dynamics of the pneumococcal population in Russia following the introduction of PCV13 immunization in 2014.

Background

Azithromycin is a useful therapeutic, but its long half-life encourages development of azolide/macrolide resistance. We investigated increasing macrolide-resistance among Streptococcus pneumoniae, exploring the association of macrolide-resistance with serotype, clone and azithromycin use.

Methods

Between January-2002 to March-2015, 464 invasive pneumococcal isolates were collected and whole-genome sequenced. Azolide/macrolide non-susceptibility was determined by erythromycin disk-diffusion and E-test.

Results

We identified macrolide resistance in 70 (15%) pneumococci; 64% of which harbored ermB, 33% mefA and 1.4% carried both the genetic determinants. Few (n=6 of 265 tested) were identified through 2009; subsequently resistance increased to 46% in 2014, and 60% in 2015. Macrolide-resistant pneumococci exhibited 24 serotypes; 19F (14%), 6B (13%) and 23F (13%) were predominant. PCV10, PCV13, and PCV20 would address 51% (36/70), 56% (39/70), and 63% (44/70), respectively. Macrolide-resistant pneumococci belonged to 26 GPSCs and 42 STs. Dominant lineages were GPSC10 (ST1553, 12894, 14490, 14488), GPSC43 (ST4745, 3214), GPSC101 (ST2854, 1078) and GPSC482 (ST5612). Increased azithromycin consumption showed direct association with increasing macrolide-resistance during this period (r=0.8572, p=0.0031, Spearman correlation coefficient).

Conclusions

Serotype and genotype diversity among macrolide-resistant pneumococci and low proportion addressed by PCVs suggests that a vaccine covering all strains or restricted consumption of azithromycin is needed to reduce transmission of macrolide-resistant strains.

Background

In 2010, Brazil introduced PCV10 into the national children’s immunization program. This study describes the genomic population structure of invasive SPN before and after PCV10 introduction.

Methods

As part of Global Pneumococcal Sequencing (GPS) project, 466 (pre-PCV10:n=232, post-PCV10:n=234; <5-year-olds:n=310, ≥5-year-olds:n=156) invasive SPN collected through national laboratory surveillance were whole-genome sequenced.

Results

The study identified 65 GPS clusters (GPSCs): 49 (88%) GPS previously described and 16 (12%) were Brazilian clusters. 36 GPSCs (55%) were non-PCV10 lineages, 11 (17%) PCV10/non-PCV10 and 18 (28%) PCV10. In both periods, the most frequent lineage was GPSC6/CC156/PMEN3/14-9V. Post-vaccine non-PCV10 lineages GPSC16/CC66/9N-15A, GPSC12/CC180/PMEN3/3 and GPSC32/CC218/PMEN24/12F increased; in <5-year-olds, GPSC1/CC320/DLV-PMEN14/19A, GPSC47/CC386/DLV-PMEN20/6C and GPSC51/CC458/3; and ≥5-year-olds GPSC3/CC53/PMEN33/8 were predominant (Figure-1).

SPN penicillin nonsusceptibility was predicted in 40%; 127 PBP combinations were identified (51 predicted MIC≥0.125mg/L); cotrimoxazole (folA+folP alterations), macrolide (mef/ermB/ermB+mef) and tetracycline (tetM/tetO/tetS/M) resistance were predicted in 46%, 13% and 21% SPN, respectively. In <5-year-olds, a penicillin (p=0.0169) and cotrimoxazole (p<0.0001) resistance reduction and an increase in tetracycline (p=0.019) were observed. Post-PCV10, PBP15-12-18(2mg/L) was frequent in lineage GPSC6/CC156/PMEN3/14-9V; among <5-year-olds the PBP13-11-16(4mg/L) in GPSC1/CC320/DLV-PMEN14/19A and PBP2-53-77(0.125mg/L) in GPSC47/ST386/DLV-PMEN20/6C were predominant.

Conclusions

Post-PCV10, important non-PCV10 lineages, GPSC1/CC320/DLV-PMEN14/19A and GPSC47/ST386/DLV-PMEN20/6C associated with multidrug resistance and GPSC12/CC180/PMEN31/3, GPSC3/CC53/PMEN33/8 and GPSC32/CC218/PMEN24/12F were identified in Brazil.