Zhongya Li, United States of America

Centres for Disease Control and Prevention Bacterial diseases

Author Of 2 Presentations

BACTERIAL MUTATIONS ASSOCIATED WITH MENINGITIS AMONG INVASIVE PNEUMOCOCCAL DISEASE (IPD) PATIENTS (ID 826)

Abstract

Background

Our understanding on bacterial genetic determinants of pneumococcal disease manifestation is still limited. We aim to confirm the previous findings and identify additional bacterial variants associated with meningitis.

Methods

We sequenced IPD isolates identified through the Active Bacterial Core surveillance (ABCs) in the United States from 2016 to 2017. We evaluated the association between meningitis and a previously reported pneumococcal pbp1b641C allele by using a mixed-effects logistic regression model accounting for population structure (represented by multi-locus sequence type) and potential confounders (pneumococcal serotype, antibiotic resistance, and patient age). We also performed a k-mer based bacterial genome-wide association study (GWAS).

Results

Among all 5560 sequenced IPD isolates, 371 (6.7%) were from meningitis cases. Among the 576 isolates carrying the pbp1b641C allele, 86 (14.9%) were meningitis. After adjusting for covariates, the pbp1b641C allele was significantly associated with meningitis (OR=1.76, 95% CI,1.10-2.81). Pneumococcal genome contents explained 8.2% (95% CI,1.6%-13.2%) of variation in meningitis in the GWAS. Additional pneumococcal mutations that showed significant association were identified in loci including SP_1448 (conserved hypothetical protein, p=1.7×10-8), SP_2167 (L-fuculokinase, p=3.6×10-8), and SP_0647 (phosphotransferase, p=1.0×10-7).

Conclusions

IPD manifestation varied significantly according to pneumococcal genomes. More knowledge on such mutations could help better understand bacterial pathogenesis and clinical outcome.

Hide

GENOMIC CLUSTERS OF INVASIVE PNEUMOCOCCAL DISEASE (IPD) ISOLATES, USA, 2015-2017 (ID 833)

Abstract

Background

Outbreaks of pneumococcal infections have been reported in closed settings and are often caused by clusters of genomically highly related isolates (genomic clusters), suggesting close transmission connections.

Methods

We used whole-genome sequencing (WGS) to characterize all IPD isolates identified through the Active Bacterial Core surveillance (ABCs) from 2015 to 2017. We identified genomic clusters by performing hierarchical cluster analysis of pair-wise genomic distances and applying a cut-off value of 3-base difference per 1.8Mb shared genome.

Results

WGS characterized 8029 isolates representing 87% of all IPD cases. The cluster analysis identified 379 genomic clusters accounting for 847 (11%) of the isolates. Higher proportions of clustered isolates (isolates belonging to any clusters) were found in serotypes 12F (36%), 4 (26%), and 7F (25%), while lower proportions were found in serotypes 35F, 7C, 15A, 23B (each <1.5%), and penicillin non-susceptible isolates (4%). Clustered isolates were significantly associated with patients who were homeless (OR=3.3), who were injecting drugs (OR=2.1), and who were 25 to 50 years of age (OR=1.5; all p-values < 0.001).

Conclusions

IPD genomic clusters were associated with specific patient and strain features in the ABCs population. Understanding the dynamics and demographics of vaccine-serotype clusters could help identify new target groups to inform vaccine policy.

Hide