Multistate population and whole genome sequence-based strain surveillance of invasive pneumococci recovered in the USA during 2017

Abstract

Objectives: We aimed to provide population-based and whole-genome sequence (WGS) -based characterization of invasive pneumococcal disease isolates collected from multistate surveillance in the USA during 2017.

Methods: We obtained short-read WGS from 2881 isolates with associated bioinformatics pipeline strain feature predictions. For quality control, capsular serotypes and antimicrobial MICs were also obtained conventionally from 442 isolates. Annotated WGS were provided (inclusive of serotypes, MICs, multilocus sequence types, pilus type(s)) from 2723 isolates. For 158 isolates with suboptimal WGS, antimicrobial MICs were obtained conventionally.

Results: There were 127 isolates from children <5 years of age and 2754 isolates from those ≥5 years old in 2017. One of 43 different serotypes was predicted for 2877 of the 2881 isolates. Serotypes in the 13-valent conjugate vaccine together with 6C (PCV13+6C) accounted for 816 (28.3%) isolates, with PCV13 serotype 3 being the most common serotype overall. Non-PCV13-6C- serotypes accounted for 2065 (71.7%) isolates, comprising 96 (75.6%) isolates from children < 5 years old and 1969 (61.4%) isolates from those aged ≥5 years. Of 36 different categories of recently emerged serotype-switch variants, three showed marked increases relative to 2015-2016 in that the number from 2017 surpassed the number from 2015-2016 combined. Two of these included antimicrobial-resistant serotype 11A and 35B serotype-switch variants of the ST156 clonal complex.

Conclusions: PCV13+6C strains are still identified in 2017 but non-PCV13-type strains impose a considerable burden. This well-annotated year 2017 WGS/strain data set will prove useful for a broad variety of analyses and improved our understanding of invasive pneumococcal disease-causing strains in the post-PCV13 era.

Keywords: Capsular serotypes; Clonal complexes; Invasive pneumococcal disease incidence; Resistance features; Serotype switch variants.

Fig. 1.

Active Bacterial Core surveillance (ABCs) invasive pneumococcal disease rates for four different age groups during 2017.

Fig. 2.

Active Bacterial Core surveillance (ABCs) serotype distributions among individuals ≥5 years (top) and <5 years (bottom). Serotypes on both graphs are in order of overall incidence. Serotypes in red font are PCV13 serotypes and 6C. Serotypes below the x-axis on lower graph were not found in the <5-year group. The primary clonal complex (green) indicated for each serotype corresponds to the single most commonly occurring clonal complex which is listed for each isolate in the Supplementary material (Table S1).

Fig. 3.

(a) Serotype-specific profiles of individual and combined resistance phenotypes for penicillin, erythromycin and clindamycin. Serotypes are listed in order of incidence. (b) Increased proportions of erythromycin-resistance and combined erythromycin/clindamycin-resistance associated with increasing penicillin MICs. The numbers of isolates within each category are also indicated. (c) Serotype associations with pilus type(s). (d) Proportions of pilus 1-positive strains increase with increasing penicillin MICs.

Fig. 3.

(a) Serotype-specific profiles of individual and combined resistance phenotypes for penicillin, erythromycin and clindamycin. Serotypes are listed in order of incidence. (b) Increased proportions of erythromycin-resistance and combined erythromycin/clindamycin-resistance associated with increasing penicillin MICs. The numbers of isolates within each category are also indicated. (c) Serotype associations with pilus type(s). (d) Proportions of pilus 1-positive strains increase with increasing penicillin MICs.

Fig. 4.

(a) Phylogeny of year 2017 serotype 4 Active Bacterial Core surveillance (ABCs) isolates depicting two unrelated and geographically segregated clonal complexes. The evolutionary history of serotype 4 isolates was inferred using KSNP3.0 to construct core single nucleotide polymorphism (SNP) matrix. The analysis involved 61 of the 65 serotype 4 genomic sequences (4 with >200 assembly contigs were not used). The tree is drawn to scale using the Mega 7 program [17], with branch lengths measured in the number of substitutions per site. There were a total of 9656 positions in the final data set by using the Maximum Likelihood method based on the General Time Reversible model. The tree with the highest log likelihood (−36461.45) is shown. Initial tree(s) for the heuristic search were obtained automatically by applying the Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood approach, and then selecting the topology with superior log likelihood value. (b) Expansion of the CC439 portion of (b) to depict relatedness between individual isolates.