PDF
Citation
Print
INTRODUCTION
Bacteremia is a major cause of morbidity and mortality in patients who receive inadequate antimicrobial treatment [1]. The case fatality rate for bacteremia is 30–40% [2]. In Korea, bacteremia caused by major antimicrobial-resistant pathogens, especially among patients hospitalized in intensive care units, has a high incidence [3, 4]. Furthermore, extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae, vancomycin-resistant Enterococcus faecium, and imipenem-resistant Acinetobacter baumannii have been on the rise [5-7].
Rapid antimicrobial susceptibility testing (AST) results are important for the selection of suitable anti-bacterial treatments for bacteremia. The main problem with current AST methods is the long turnaround time (TAT). In most cases, conducting AST requires overnight incubation and usually requires 48–72 hours to complete, depending on the drug–organism combination [8]. Conventional AST devices such as Vitek 2 (bioMérieux Inc., Marcy l’Etoile, France), MicroScan WalkAway (Beckman Coulter Inc., Carlsbad, CA, USA), and BD Phoenix (BD Diagnostic Systems, Sparks, MD, USA) are now widely used in clinical laboratories, but these systems take at least 8 to 20 hours to produce AST results, not including bacterial incubation and isolation time [9].
Here, we describe a direct & Rapid Antimicrobial Susceptibility Testing (dRAST) system called QMAC-dRAST (QuantaMatrix Inc., Seoul, Republic of Korea) which can shorten TAT to less than 30 hours. This system can process a wide dynamic range of inoculum sizes directly from positive blood culture bottles without measuring inoculum size and without a separation process [10]. QMAC-dRAST is based on a microfluidic system using plastic microchips consisting of micropatterned radial chambers containing an agarose matrix with patient blood samples and satellite wells containing freeze-dried antibiotics at different concentrations. Automatic time-lapse microscopic imaging technology of this system analyzes bacterial microcolony growth in 6 hours directly in the wells [10-13]. Several studies have evaluated the clinical performance of QMAC-dRAST [14-16]. However, previous studies did not show agreement rates classified by bacterial species between other AST systems and QMAC-dRAST. In this study, we evaluated QMAC-dRAST using positive blood culture samples and presented data classified by antimicrobial agents or bacterial species. Also, we checked for changes in AST results made by an updated interpretation algorithm of AST results (AST algorithm).
MATERIALS AND METHODS
1. Specimen collection
Specimens were collected from October 2017 to March 2018 from positive blood culture samples of patients at Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea. This study was approved by the Severance Hospital Institutional Review Board, Seoul, Korea (IRB No.1-2017-0079). The total number of isolates was 204, with 104 Gram-positive cocci (GPC) and 100 Gram-negative rods (GNR), including Staphylococcus spp., Enterococcus spp., Enterobacteriaceae, Pseudomonas aeruginosa, and Acinetobacter spp. Polymicrobial blood culture samples were excluded from this study. All isolates were identified using the Vitek 2 and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry identification system (Bruker Daltonics Inc., Billerica, MA, USA). These isolates, which had been inoculated in Mueller Hinton broth containing 15% glycerol and stored at -70°C, were cultured in blood agar.
2. Bacteria spiking protocol
For AST using the QMAC-dRAST method, each bacterial colony from blood agar plates was mixed with 1.0 mL of 0.9% saline in a glass tube. Then, a 0.5 McFarland Standard suspension of the tested bacteria was made (approximately 1.5×108 CFU/mL). Serial dilution was performed, resulting in a final bacterial concentration of approximately 1.0×103 CFU/mL, and 1.0 mL of the final diluted sample was inoculated into a Bact/Alert FA blood culture bottle containing 5.0 mL of sheep blood. A sample from the bottle was taken after the Bact/Alert 3D system (bioMérieux Inc., Marcy l’Etoile, France) detected it as positive and was then tested directly in the QMAC-dRAST system. After the first test with 67 GPC and 72 GNR, the AST algorithm was updated. We reexamined 67 GPC and 72 GNR using the new AST algorithm in the second test. Then, an additional 37 GPC and 28 GNR were tested in the final test.
3. Performance evaluation of the QMAC-dRAST system
We compared the results of QMAC-dRAST with those of Vitek 2. For discrepant AST results between QMAC-dRAST and Vitek 2, broth microdilution (BMD) tests were additionally conducted to resolve discrepancies. Different types of antibiotics were tested for each of the bacteria according to the QMAC-dRAST product panel. The Vitek 2 ASTs were conducted according to the manufacturer’s guidelines and BMD tests were performed based on the Clinical and Laboratory Standards Institute (CLSI) guidelines [17].
4. Quality control
For quality control, three GPC strains (S. aureus ATCC 29213, Enterococcus faecalis ATCC 29212, E. faecalis ATCC 51299) and three GNR strains (Escherichia coli ATCC 25922, P. aeruginosa ATCC 27853, K. pneumoniae ATCC 700603) were used for the QMAC-dRAST system and BMD tests. For the Vitek 2 system, three GPC strains (S. aureus ATCC 29213, E. faecalis ATCC 29212, E. faecalis ATCC 51299) and two GNR strains (E. coli ATCC 25922 and P. aeruginosa ATCC 27853) were used.
5. Data analyses
The isolates were categorized into resistant (R), susceptible (S), or intermediate (I) groups based on the AST results of Vitek 2 according to the manufacturer’s instructions. The concordance of results between QMAC-dRAST and Vitek 2 was determined using categorical agreement (CA, i.e., agreement of results between the test method and the comparator) and essential agreement (EA, i.e., agreement within±1 two-fold dilution of the test method under evaluation with the comparator minimum inhibitory concentration determination). Discrepant results for the detection of antimicrobial susceptibility were classified as very major errors (VME, i.e., susceptible by the test method vs. resistant by the comparator), major errors (ME, i.e., resistant by the test method vs. susceptible by the comparator), and minor errors (mE, i.e., intermediate by the test method vs. resistant or susceptible by the comparator or vice versa) [18]. Data are presented as numbers with percentages for categorical variables.
RESULTS
1. Evaluation of QMAC-dRAST algorithm improvement
In the first test, 67 GPC and 72 GNR were tested and the agreement rates between QMAC-dRAST and Vitek 2 results were analyzed. Discrepant results were found in certain AST cases. Therefore, we updated the AST algorithm. In the second test, ASTs of 67 GPC and 72 GNR were performed again by QMAC-dRAST after AST algorithm improvement. We compared AST results between QMAC-dRAST and Vitek 2 before and after AST algorithm optimization. As a result of the AST algorithm update, tetracycline susceptibility test results for 9 isolates of Staphylococcus spp. (4 S. epidermidis, 1 S. haemolyticus, 2 S. hominis, 1 S. warneri, and 1 coagulase-negative staphylococcus [CoNS]) were corrected. A ME occurred in oxacillin AST results for S. capitis and three MEs in the results for imipenem (1 Enterobacter cloacae, 1 E. coli, 1 K. pneumoniae) were also corrected. For GPC, AST algorithm improvement increased CA from 92.7% to 94.3%, and decreased VME from 6.5% to 5.1% and ME from 2.7% to 2.4%; for GNR, it increased CA from 96.2% to 96.5%, had no effect on VME (1.2%), and decre-ased ME from 1.6% to 1.2% (Table 1).
2. Performance evaluation of QMAC-dRAST for all samples
In the final test, an additional 37 GPC and 28 GNR were included, so that 104 GPC and 100 GNR were tested after AST algorithm improvement. A total of 913 AST results for QMAC-dRAST with GPC and 1,314 results with GNR were compared with Vitek 2 results. For GPC, the agreement rates were 94.5% for CA, 4.8% for VME, 2.5% for ME, and 2.1% for mE. For GNR, the agreement rates were 95.4% for CA, 0.6% for VME, 2.0% for ME, and 3.0% for mE. Most VMEs in GPC were detected for S. aureus and CoNS (Table 2). In vancomycin susceptibility tests, among 17 GPC isolates determined to be resistant, only one (5.9%) E. faecium isolate showed a false susceptible result by QMAC-dRAST (Table 3). GNR samples yielded VMEs only in E. coli but produced relatively many MEs in some species (Table 2). In GPC, VMEs were found for most of the tested antibiotics except for ampicillin, gentamicin, and linezolid; most MEs were discovered for penicillin and cotrimoxazole. However, in GNR, only ceftazidime and piperacillin/tazobactam produced VMEs, but most antibiotics yielded MEs (Table 3).
DISCUSSION
Presently, four devices are capable of automated AST: Vitek 2, MicroScan WalkAway, BD Phoenix, and Sensititre ARIS 2X (TREK Diagnostic Systems Inc., Westlake, OH, USA). The first three systems generate results within 3.5–16 hours, whereas Sensititre ARIS 2X takes more time on average to report the final results [19]. However, those three faster methods require standardized microbial inoculation and also have the disadvantage of not being able to conduct AST directly with positive blood culture bottles. Samples should be cultured for 24–48 hours or longer before inoculation into the AST system [19-21].
However, QMAC-dRAST reports AST results only by observing changes in bacterial cell shape using microscopic imaging without the need for incubation [10, 12, 13]. Thus, TAT can be reduced by about 24 to 48 hours, reducing the empirical treatment period, which is especially important for major antibiotic-resistant pathogens such as carbapenem-resistant Enterobacteriaceae, vancomycin-resistant enterococci, and methicillin-resistant S. aureus [22-25]. Empirical treatment is often unsuitable for antimicrobial-resistant bacteremia, and the quick delivery of results helps clinicians reduce empirical anti-bacterial drug administration and respond to antimicrobial resistance.
In this study, an older QMAC-dRAST system (software v.1.0.13) was updated to resolve discrepancies. In the new QMAC-dRAST system (software v.1.0.14), the AST algorithm was updated for the following antimicrobial agent-bacterial species combinations: tetracycline-Staphylococcus spp., trimethoprim/sulfamethoxazole-Staphylococcus spp., and imipenem-GNR.
In the case of GPC, the overall performance of QMAC-dRAST met the criteria for AST in the Food and Drug Administration (FDA) guidelines (e.g. CA ≥90%, VME ≤1.5%, ME ≤3%) [18], except for VME (Table 2). Most VMEs were detected in S. aureus and CoNS. Therefore, it is necessary to improve the performance of QMAC-dRAST for staphylococci. In a previous study, Huh et al. evaluated the performance of QMAC-dRAST for staphylococci [15]. However, since agreement rates for staphylococci were presented together with enterococci in that study, a direct comparison with our study was not possible.
The overall performance of QMAC-dRAST with GNR met the FDA guidelines described above. Although we detected a small number of VMEs in E. coli, there was no VME for commonly used antimicrobial agents such as β-lactam/β-lactamase inhibitor, broad-spectrum cephalosporins, and carbapenem, except for ceftazidime and piperacillin/tazobactam in the case of GNR (Tables 2, 3).
Our study had certain limitations. First, BMD tests were performed only for samples that showed discrepancies between QMAC-dRAST and Vitek 2. In further studies, comparing all tests with BMD as a reference method will be desirable. Second, clinically rare bacterial strains were not included in this study. Therefore, the results of this study may not be applied to some other bacterial strains. More bacterial strains may be required to validate our results. However, the main strength of our study is that we presented the results classified by bacterial species.
In conclusion, the QMAC-dRAST automated system was comparable to Vitek 2 and had the advantage of reporting AST results more rapidly. The concordance of this system with conventional devices is likely to be further improved in the future by updating AST algorithms.
XML Download