Genomic surveillance of Acinetobacter baumannii in the Philippines, 2013–2014

Abstract

Objective: Acinetobacter baumannii is an opportunistic nosocomial pathogen that has increasingly become resistant to carbapenems worldwide. In the Philippines, rates of carbapenem resistance and multidrug resistance are above 50%. We undertook a genomic study of carbapenem-resistant A. baumannii in the Philippines to characterize the population diversity and antimicrobial resistance mechanisms. Methods: We sequenced the whole genomes of 117 A. baumannii isolates recovered by 16 hospitals in the Philippines between 2013 and 2014. From the genome sequences, we determined the multilocus sequence type, presence of acquired determinants of antimicrobial resistance and relatedness between isolates. We also compared the phenotypic and genotypic resistance results. Results: Carbapenem resistance was mainly explained by acquisition of the class-D Beta-lactamase gene blaOXA-23. The concordance between phenotypic and genotypic resistance to imipenem was 98.15%, and it was 94.97% overall for the seven antibiotics analysed. Twenty-two different sequence types were identified, including 7 novel types. The population was dominated by the high-risk international clone 2 (i.e. clonal complex 92), in particular by ST195 and ST208 and their single locus variants. Using whole-genome sequencing, we identified local clusters representing potentially undetected nosocomial outbreaks, as well as multi-hospital clusters that indicated interhospital dissemination. Comparison with global genomes suggested that the establishment of carbapenem-resistant international clone 2 in the Philippines is likely the result of clonal expansion and geographical dissemination, and at least partly explained by inadequate hospital infection control and prevention. Discussion: This is the first extensive genomic study of carbapenem-resistant A. baumannii in the Philippines, and it underscores the importance of hospital infection control and prevention measures to contain high-risk clones.

Genomic Characterization of Carbapenem-Resistant Acinetobacter baumannii (CRAB) in Mechanically Ventilated COVID-19 Patients and Impact of Infection Control Measures on Reducing CRAB Circulation during the Second Wave of the SARS-CoV-2 Pandemic in Milan, Italy

“…To define potential transmission chains, we decided to apply an SNP threshold of <10 among isolates. This threshold is in line with recent published papers ( 24 , 30 ), even if a little above the 2.5 SNPs proposed by Coll et al ( 23 ). Regarding this point, we need to highlight that the transmission chains here described suffered the introduction of CRAB isolates originating in different hospitals of the Lombardy region and had quite different durations.…”

“…In accordance with the genetic relatedness of the 40 A. baumannii strains, a threshold of 10 SNPs was considered suggestive of potential transmission chains. This threshold is consistent with already published articles on bacterial divergence, which consider a minimum of 2.5 and maximum of 15 core genome SNPs to rule out transmission chains ( 22 – 24 ).…”

“…It is noteworthy that our work reinforced the idea that an integration of methodologies investigating both the core genome, through ML, MST, and SNP analysis, and the whole genome, by identification of AMR and virulence determinants, as well as mobile genetic elements, is essential to correctly identify and characterize potential transmission chains ( 22 – 24 ). Indeed, the core genome SNP approach alone could fail to detect essential information like loss or acquisition of entire genes, while whole-genome sequencing could be difficult to apply to maximum likelihood or Bayesian methods, especially when a huge number of isolates needs to be studied.…”

“…The ML analysis confirmed the clustering of the strains based on the five Oxford STs and allowed identification of differences in AMR, virulence, and MGEs among and within the different STs. Based on the ML and minimum spanning trees obtained by the core genome and through the definition of a threshold of 10 core SNPs (a threshold concordant with the literature in references 22 to 24 ), we were able to identify four putative transmission chains, involving over two-thirds of CRAB strains (30/40). The genetic relatedness of these strains was also confirmed by the fact that they almost always shared the same AMR and virulence genes and mobile genetic elements.…”

“…From the results obtained by the ML tree, minimum spanning tree, SNP distance, and whole-genomic analysis, it was evident the presence of 4 putative transmission chains, constituted by ≥3 strains with an intracluster pairwise SNP distance always lower than 10 ( 22 – 24 ). Hence, a Bayesian analysis was performed to better characterize these putative transmission chains, after the removal of a total of 10 sequences (IDs 1363, 1423, 1403, 1495, 1482, 1353, 1399, 1400, 1450, and 1506), differing for a number of core SNPs of ≥10 compared to the strains of the same ML cluster.…”