Ana Ong

The Challenges of Implementing Next Generation Sequencing Across a Large Healthcare System, and the Molecular Epidemiology and Antibiotic Susceptibilities of Carbapenemase-Producing Bacteria in the Healthcare System of the U.S. Department of Defense

Objective We sought to: 1) provide an overview of the genomic epidemiology of an extensive collection of carbapenemase-producing bacteria (CPB) collected in the U.S. Department of Defense health system; 2) increase awareness of the public availability of the sequences, isolates, and customized antimicrobial resistance database of that system; and 3) illustrate challenges and offer mitigations for implementing next generation sequencing (NGS) across large health systems. Design Prospective surveillance and system-wide implementation of NGS. Setting 288-hospital healthcare network. Methods All phenotypically carbapenem resistant bacteria underwent CarbaNP® testing and PCR, followed by NGS. Commercial (Newbler and Geneious), on-line (ResFinder), and open-source software (Btrim, FLASh, Bowtie2, an Samtools) were used for assembly, SNP detection and clustering. Laboratory capacity, throughput, and response time were assessed. Results From 2009 through 2015, 27,000 multidrug-resistant Gram-negative isolates were submitted. 225 contained carbapenemase-encoding genes (most commonly blaKPC, blaNDM, and blaOXA23). These were found in 15 species from 146 inpatients in 19 facilities. Genetically related CPB were found in more than one hospital. Other clusters or outbreaks were not clonal and involved genetically related plasmids, while some involved several unrelated plasmids. Relatedness depended on the clustering algorithm used. Transmission patterns of plasmids and other mobile genetic elements could not be determined without ultra-long read, single-molecule real-time sequencing. 80% of carbapenem-resistant phenotypes retained susceptibility to aminoglycosides, and 70% retained susceptibility to fluoroquinolones. However, among the CPB-confirmed genotypes, fewer than 25% retained susceptibility to aminoglycosides or fluoroquinolones. Conclusion Although NGS is increasingly acclaimed to revolutionize clinical practice, resource-constrained environments, large or geographically dispersed healthcare networks, and military or government-funded public health laboratories are likely to encounter constraints and challenges as they implement NGS across their health systems. These include lack of standardized definitions and quality control metrics, limitations of short-read sequencing, insufficient bandwidth, and the current limited availability of very expensive and scarcely available sequencing platforms. Possible solutions and mitigations are also proposed.

Correlating Cleaning Thoroughness with Effectiveness and Briefly Intervening to Affect Cleaning Outcomes: How Clean Is Cleaned?

Objectives The most efficient approach to monitoring and improving cleaning outcomes remains unresolved. We sought to extend the findings of a previous study by determining whether cleaning thoroughness (dye removal) correlates with cleaning efficacy (absence of molecular or cultivable biomaterial) and whether one brief educational intervention improves cleaning outcomes. Design Before-after trial. Setting Newly built community hospital. Intervention 90 minute training refresher with surface-specific performance results. Methods Dye removal, measured by fluorescence, and biomaterial removal and acquisition, measured with culture and culture-independent PCR-based assays, were clandestinely assessed for eight consecutive months. At this midpoint, results were presented to the cleaning staff (intervention) and assessments continued for another eight consecutive months. Results 1273 surfaces were sampled before and after terminal room cleaning. In the short-term, dye removal increased from 40.3% to 50.0% (not significant). For the entire study period, dye removal also improved but not significantly. After the intervention, the number of rooms testing positive for specific pathogenic species by culturing decreased from 55.6% to 36.6% (not significant), and those testing positive by PCR fell from 80.6% to 53.7% (P = 0.016). For nonspecific biomaterial on surfaces: a) removal of cultivable Gram-negatives (GN) trended toward improvement (P = 0.056); b) removal of any cultivable growth was unchanged but acquisition (detection of biomaterial on post-cleaned surfaces that were contaminant-free before cleaning) worsened (P = 0.017); c) removal of PCR-based detection of bacterial DNA improved (P = 0.046), but acquisition worsened (P = 0.003); d) cleaning thoroughness and efficacy were not correlated. Conclusion At this facility, a minor intervention or minimally more aggressive cleaning may reduce pathogen-specific contamination, but not without unintended consequences.

Relationships among cleaning, environmental DNA, and healthcare-associated infections in a new evidence-based design hospital.

OBJECTIVEnHospital environments influence healthcare-associated infection (HAI) patterns, but the role of evidenced-based design (EBD) and residual bacterial DNA (previously thought to be clinically inert) remain incompletely understood.nnnMETHODSnIn a newly built EBD hospital, we used culture-based and culture-free (molecular) assays, pulsed-field gel electrophoresis (PFGE), and whole-genome sequencing (WGS) to determine: (1) patterns of environmental contamination with target organisms (TOs) and multidrug-resistant (MDR) target organisms (MDR-TOs); (2) genetic relatedness between environmentally isolated MDR-TO and those from HAIs; and (3) correlation between surface contamination and HAIs.nnnRESULTSnA total of 1,273 high-touch surfaces were swabbed before and after terminal cleaning during 77 room visits. Of the 2,546 paired swabs, 47% had cultivable biomaterial and 42% had PCR-amplifiable DNA. The ratios of TOs detected to surfaces assayed were 85 per 1,273 for the culture-based method and 106 per 1,273 for the PCR-based method. Sinks, toilet rails, and bedside tables most frequently harbored biomaterial. Although cleaned surfaces were less likely to have cultivable TOs than precleaned surfaces, they were not less likely to harbor bacterial DNA. The rate of MDR-TOs to surfaces swabbed was 0.1% (3/2546). Although environmental MDR-TOs and MDR-TOs from HAIs were genetically related by PFGE, WGS revealed that they were unrelated. Environmental levels of cultivable Enterococcus spp. and E. coli DNA were positively correlated with infection incidences (P<.04 and P<.005, respectively).nnnCONCLUSIONnMDR-TOs were rarely detected during surveillance and were not implicated in HAIs. The roles of environmental DNA and EBD, particularly with respect to water-associated fixtures or the potential suppression of cultivable environmental MDR-TOs, warrant multicenter investigations.

Got GAS? Ease the Bloat with Real-Time Whole-Genome Sequencing

Financial support: Support for this study was received from the German Research Foundation (grant no.WO 1746/1-2 toM.W. and grant no. KA 4199/ 1-1 to T.H.). Additional funding was received from the Innovative Medicines Initiative Joint Undertaking (grant no. 115737-2 to K.K., Combatting bacterial resistance in Europe—molecules against gram-negative infections [COMBACTE-MAGNET]). The funders had no role in the preparation of this manuscript. Potential conflicts of interest: All authors report no conflicts of interest relevant to this article.

From the Battlefield to the Bedside: Supporting Warfighter and Civilian Health With the “ART” of Whole Genome Sequencing for Antibiotic Resistance and Outbreak Investigations

Awareness, responsiveness, and throughput characterize an approach for enhancing the clinical impact of whole genome sequencing for austere environments and for large geographically dispersed health systems. This Department of Defense approach is informing interagency efforts linking antibiograms of multidrug-resistant organisms to their genome sequences in a public database.

Clinical and Genomic Features of the First Cases of Elizabethkingia anophelis Infection in New York, Including the First Case in a Healthy Infant Without Previous Nosocomial Exposure

Elizabethkingia spp are Gram-negative bacteria associated with neonatal meningitis. In 2015-2016, an outbreak of Elizabethkingia anophelis infection that involved 63 patients and 18 deaths occurred in Wisconsin. Despite a multistate investigation, as of September 2016 the source remained undetermined, and experts warned of reemergence. We describe here the first cases of E anophelis infection in New York, including the case of a healthy infant without previous healthcare contact.

Pseudomona s Endocarditis with an unstable phenotype: the challenges of isolate characterization and Carbapenem stewardship with a partial review of the literature

BackgroundPseudomonas endocarditis is exceedingly rare, especially in patients without predisposing risks. We present such a case that included unexpected switches in antibacterial resistance profiles in two Pseudomonas aeruginosa (PA) strains with the same whole-genome sequence. The case also involved diagnostic and treatment challenges, such as issues with automated testing platforms, choosing the optimal aminoglycoside, minimizing unnecessary carbapenem exposure, and the need for faster, more informative laboratory tests.Case presentationOn hospital day one (HD-1) a cefepime and piperacillin-tazobactam (FEP-TZP)-susceptible P. aeruginosa was isolated from the bloodstream of a 62-year-old man admitted for evaluation of possible endocarditis and treated with gentamicin and cefepime. On HD-2, his antibiotic regimen was changed to tobramycin and cefepime. On HD-11, he underwent aortic valve replacement, and P. aeruginosa was isolated from the explanted valve. Unexpectedly, it was FEP-TZP-resistant, so cefepime was switched to meropenem. On HD-14, in preparation for whole-genome sequencing (WGS), valve and blood isolates were removed from cryo-storage, re-cultured, and simultaneously tested with the same platforms, reagents, and inoculations previously used. Curiously, the valve isolate was now FEP-TZP-susceptible. WGS revealed that both isolates were phylogenetically identical, differing by a single nucleotide in a chemotaxis-encoding gene. They also contained the same resistance genes (blaADC35, aph(3′)-II, blaOXA-50, catB7, fosA).ConclusionRepeated testing on alternate platforms and WGS did not definitively determine the resistance mechanism(s), which in this case, is most likely unstable de-repression of a chromosomal AmpC β-lactamase, porin alterations, or efflux upregulation, with reversion to baseline (non-efflux) transcription. Although sub-culture on specialized media to select for less fit (more resistant) colonies, followed by transcriptome analysis, and multiple sequence alignment, might have revealed the mechanism and better informed the optimal choice of β-lactam, such approaches are neither rapid, nor feasible for hospital laboratories. In this era of escalating drug resistance and dwindling antibiotics, use of the most potent anti-pseudomonals must be balanced with stewardship. Clinicians need access to validated genomic correlates of resistance, and faster, more informative diagnostics. Therefore, we placed these isolates and their sequences in the public domain for inclusion in the Pseudomonas pan-genome and database projects for further countermeasure development.

1671A Fatal Outbreak of a Rare but Emerging Clone of Extensively Drug-resistant Acinetobacter baumannii with Enhanced Virulence

Background. The virulence mechanisms of Acinetobacter baumannii(AB) are not completely understood. Also, a high degree of virulence in immunocompetent hosts is unusual for AB. A cluster of six fatal AB infections in mostly immune-competent patients who received early infectious diseases consultation and appropriate antimicrobial therapy led us to hypothesize that isolates from these patients could possess unique traits that enhance virulence. Methods. To test our hypothesis, we determined the health status of patients using three different co-morbidity and illness severity scoring systems, characterized isolates using comparative genomics, and evaluated the virulence of each isolate in a validated animal model. Results. No patient had scores that indicated excessive co-morbidities or critical physiologic dysfunction. Two patients had bacteremia scores that indicated severe illness. Genomic analysis showed that two unrelated AB clones were associated with the infections. The clone associated with the majority (5 of 6) of patient deaths, clade B, is evolutionarily distinct from the three main international clonal complexes and is most closely related to strains isolated in the Czech Republic in 1994 and in a U.S. military hospital in Germany in 2003. One clade B isolate, MRSN16897, is hyper-virulent in a murine pulmonary model of infection when compared to other well-known and previously tested AB isolates. Of note, MRSN16987 establishes disseminating, fatal infections, even in healthy mice when other strains of AB cannot. Clade B strains contain virulence genes whose products differ from those in the majority of AB strains, including loci involved in iron metabolism, protein secretion, and glycosylation. Using genome sequences of the fatal strains and related isolates from our repository, we developed a PCR-based assay to rapidly and specifically detect clade B isolates. (This clade is not distinguishable by multi-locus sequence typing.) Conclusion. We identified genomic correlates of virulence (a unique gene set in the progenitor, and 4 single nucleotide changes a hypervirulent derivitave ) in the emerging AB clone. This clone is notable for a combination of resistance and higher virulence. Our findings highlight the value of combining surveillance, biobanking, and basic research. Disclosures. All authors: No reported disclosures.