Performance of carbapenemase screening algorithms for detection of OXA-244-producing Escherichia coli

Abstract

Objectives

OXA-244-producing Escherichia coli emerge in Europe and are difficult to detect. We evaluated the performance of different carbapenemase-producing Enterobacterale (CPE) screening and confirmation protocols used across five clinical microbiology laboratories in the Netherlands for the detection of OXA-244-producing E. coli isolates.

Methods

Twelve E. coli isolates (CPE n = 9, non-CPE-ESBL n = 3) containing bla_OXA-244 (n = 5), bla_OXA-48(n = 1), bla_OXA-181 (n = 1), bla_KPC-2 (n = 1) and bla_NDM-5 (n = 1) were distributed to five laboratories. Laboratories performed identification, susceptibility tests and subsequent additional tests for resistance mechanisms in accordance with their local CPE screening and confirmation protocol which all adhered to the current Dutch guideline for CPE detection.

Results

All laboratories correctly detected no carbapenemase in three E. coli isolates without carbapenemase genes and correctly determined the E. coli isolates with bla_KPC-2 and bla_NDM-5 as CPE. Detection rates of carbapenemase among the five E. coli isolates with bla_OXA-244 ranged from 0/5 (0%) to 4/5 (80%) isolates. One E. coli isolate with bla_OXA-244, a meropenem MIC ≤0.125 mg/L and an imipenem MIC ≤0.25 mg/L was not detected by any laboratory.

Discussion

Differences in CPE screening and confirmation protocols across five laboratories contributed to inconsistent detection of OXA-244-producing E. coli. Laboratories that used a gradient strip test confirmed meropenem minimum inhibitory concentration (MIC) >0.25 mg/L CPE screening breakpoint detected fewer or no OXA-244-producing E. coli strains than laboratories that implemented additional measures in their CPE screening protocols. A key recommendation is that laboratories evaluate whether the performance of their CPE screening protocol remains adequate, considering the current epidemiology of emerging difficult-to-detect OXA-244-producing E. coli.

Introduction

Carbapenemase-producing Enterobacterales (CPE) pose a worldwide threat to human health.¹ OXA-244 carbapenemase was first described in Klebsiella pneumoniae but now predominates in Escherichia coli.² In comparison with OXA-48, OXA-244 contains a single Arg214Gln substitution, resulting in reduced carbapenem and temocillin hydrolysis, and, consequently, detection failure on some CPE screening agars.^3–5  Bla_OXA-244 in E. coli is predominantly chromosomally located onto a composite transposon.^3,6 European spread of bla_OXA-244 in E. coli was initially dominated by sequence type (ST)38.^6–8 Rapid spread of bla_OXA-244 occurs mainly by community transmission of OXA-244-producing E. coli ST131, combined with its presence in the environment and introductions into hospitals.^9–12

The ECDC found that 88.2% of OXA-244-producing E. coli was detectable by the EUCAST meropenem screening breakpoint of >0.125 mg/L.¹³ The Antibiogram Committee of the French Society of Microbiology responded to the emerging epidemiological trend of carbapenemase genes with diagnostic challenges by radically revising a disk diffusion algorithm in 2022. As a result, the sensitivity for detecting OXA-244 in Enterobacterales increased from 12/37 (32%) to 32/37 (86%), at the cost of a modest decrease in specificity.¹⁴ In contrast, in Enterobacterales, Dutch microbiology laboratories mostly perform automated susceptibility tests (AST) like Vitek-2 (BioMérieux) and BD Phoenix AST (Becton Dickinson). Since the meropenem breakpoint recommended by EUCAST for carbapenemase detection (>0.125 mg/L) is not compatible with an often-used AST card, the screening breakpoint for meropenem in the current Dutch guideline is >0.25 mg/L. CPE detection may fail more frequently in the current epidemiological situation, with upcoming, difficult-to-detect OXA-244. We aimed to assess the ability to detect OXA-244-producing E. coli isolates using different CPE screening and confirmation protocols across five Dutch clinical microbiology laboratories.

Methods

Bacterial isolates

Twelve E. coli isolates (CPE n = 9, non-CPE-ESBL n = 3) were distributed. The carbapenemase-producing E. coli isolates contained bla_OXA-244 (n = 5), bla_OXA-48(n = 1), bla_OXA-181 (n = 1), bla_KPC-2 (n = 1) and bla_NDM-5 (n = 1).

Nanopore sequencing

After genomic DNA was extracted using the QIAcube (QIAGEN), combined with the QIAGEN DNeasy Blood & Tissue kit, WGS was performed on all isolates using the Rapid Barcoding Kit 96 V14 with a GridION device (MinKNOW software v24.06.15, Dorado basecaller v7.4.14 running the super-accurate basecalling model v4.3.0 400 bps; Oxford Nanopore Technologies). Genomes were assembled and analyzed using Nanoplot,¹⁵ Porechop,¹⁶ fastp,¹⁷ Flye v2.9.3.,¹⁸ Medaka v1.7.2.¹⁹ and staramr²⁰ using the Galaxy platform,²¹ allowing the identification of resistance genes and determination of STs (Table S1, available as Supplementary data at JAC-AMR Online). Reads with a ≥Q15 and a minimal length of 500 bp were included, and an additional 30 bp was trimmed from each end after adapter removal. All but one isolates showed an L90 of ≤3 contigs and an average sequencing depth >70×. This isolate had a lower average coverage depth (23×); nevertheless, the assembly was circular, contained >95% of the cgMLST loci, and was therefore included. In addition, for all isolates, >90% of the cgMLST alleles were classified as good targets. Core genome multilocus sequence typing (cgMLST) was performed to assess clonality among isolates using an in-house-developed scheme comprising 2989 loci using Ridom SeqSphere+ software (Ridom GmbH, Münster, Germany) (Table S1, Figure S1).

Automated susceptibility testing

Laboratories 1 and 2 use BD Phoenix™ (BD Biosciences, Franklin Lakes, New Jersey, USA) AST with NMIC-408 or NMIC-475 panels, respectively. Meropenem MIC results in these panels range from ≤0.125 to >8 mg/L. Laboratories 3–5 use the VITEK^® 2 AST with AST-N447 cards (bioMérieux, Marcy-l'Étoile, France). Meropenem MIC results from these cards range from ≤0.25 to >8 mg/L.

CPE screening according to local protocol

The 12 E. coli isolates were distributed to five participating laboratories. Detailed disclosure of the CPE screening protocol of each participating laboratory was collected. Personnel from the participating laboratories were blinded to the susceptibility test results and the presence of ESBL and/or carbapenemase genes. Each laboratory performed identification, susceptibility tests and additional tests for resistance mechanisms according to their local protocol. For this study, laboratories were allowed to refrain from using WGS. Data were reported to laboratory 1. The reports included, amongst others, antibiotic susceptibility test results, and, if tested and detected, the presence of carbapenemase.

Results

Screening for carbapenemase-producing Enterobacterales

The CPE screening approach differed between the five laboratories. Initial CPE screening in laboratories 1, 3–5 was performed using automated meropenem MICs >0.25 mg/L and/or imipenem MICs >1 mg/L (Table 1). Laboratory 4 additionally included piperacillin/tazobactam MICs >8 mg/L (Table 1). Laboratory 2 uses CPO detect (a test within the BD Phoenix AST NMIC-475 panel) (Table 1). Following initial CPE screening, Etest meropenem MICs >0.25 mg/L and/or Etest imipenem MICs >1 mg/L were required in laboratories 1, 3 and 5 but not in laboratories 2 and 4. Definitive CPE confirmation by Xpert Carba-R (Cepheid) test is used in laboratories 1 and 2, carbapenem inactivation method (CIM) in laboratories 3 and 5 and the NG-TEST CARBA-5 (NG Biotech) in laboratory 4 (Table 1).

Table 1.

Laboratory algorithm for CPE detection and detection rates of CPE stratified according to participating laboratory

Laboratory	Method			Detected as CPE/tested (ndetected/ntested)
	Initial susceptibility test platform for Enterobacterales (panel/card)	Initial screening for CPE	Subsequent screening for CPE	CPE confir-mation test	E. coli OXA-244	E. coli OXA-48	E. coli OXA-181	E. coli KPC-2	E. coli NDM-5	E. coli ESBL
1	BD Phoenix (NMIC-408)	Automated meropenem MIC >0.25 mg/L and/or imipenem MIC >1 mg/L	Etest meropenem MIC >0.25 mg/L and/or Etest imipenem MIC >1 mg/L	Xpert Carba-Ra	1/5	0/1	0/1	1/1	1/1	0/3
2	BD Phoenix (NMIC-475)	Confirmatory CPO detect test (within NMIC-475 panel) and/or automated meropenem MIC >0.125 mg/L		Xpert Carba-Ra	4/5	1/1	1/1	1/1	1/1	0/3
3	Vitek 2 (AST-N447)	Automated meropenem MIC >0.25 mg/L and/or imipenem MIC >1 mg/L	Etest meropenem MIC >0.25 mg/L and/or Etest imipenem MIC >1 mg/L	CIM	0/5	0/1	0/1	1/1	1/1	0/3
4	Vitek 2 (AST-N447)	Automated piperacillin/tazobactam MIC >8.0 mg/L and/or meropenem MIC >0.25 mg/L and/or imipenem MIC >1 mg/L		NG-TEST CARBA-5bc	4/5	1/1	1/1	1/1	1/1	0/3
5	Vitek 2 (AST-N447)	Automated meropenem MIC >0.25 mg/L and/or imipenem MIC >1 mg/L	Etest meropenem MIC >0.25 mg/L and/or Etest imipenem MIC >1 mg/L	CIM	1/5	1/1	0/1	1/1	1/1	0/3

CPE, carbapenemase-producing Enterobacterale.

^aXpert Carba-R detects various blaOXA-48L, including blaOXA-244 leading to an OXA-48 detected result.

^bNG-TEST CARBA-5 detects various blaOXA-48L, including blaOXA-244 leading to a positive OXA-48L result.

^cNG-TEST CARBA-5 test is performed if the initial piperacillin/tazobactam screening is positive and/or second and/or third screening steps for CPE are positive.

Carbapenemase detection

None of the laboratories detected carbapenemases among the three E. coli isolates without carbapenemase genes (Table 1). All laboratories correctly determined the E. coli isolates with bla_KPC-2 or bla_NDM-5 as CPE (Table 1). E. coli with bla_oxa-48 was correctly determined as CPE by 3/5 laboratories and E. coli with bla_oxa-181 by 2/5 (Table 1). None of the five participating laboratories detected all OXA-244-producing E. coli isolates. Detection rates of carbapenemase among the five E. coli isolates with bla_OXA-244 ranged from 0/5 (0%) to 4/5 (80%): Laboratories 2 and 4 detected 4/5 (80%) of the E. coli isolates with bla_OXA-244, laboratories 1 and 5 both detected 1/5 (20%), and laboratory 3 detected 0/5 (0%) (Table 1). None of the laboratories detected one E. coli isolate with bla_OXA-244, a meropenem MIC ≤0.125 mg/L and an imipenem MIC ≤0.25 mg/L by BD Phoenix AST.

Bla_OXA-244 detection in relation to screening and confirmation protocol

The two laboratories with the highest performance detected 4/5 (80%) bla_OXA-244  E. coli isolates (Table 1). One used CPO detect in combination with a meropenem screening breakpoint of MIC >0.125 mg/L in BD Phoenix AST (Table 1). The other performed the NG-Test carba in isolates with a piperacillin/tazobactam MIC >8/4 mg/L without Etest confirmation, independent of but in addition to carbapenem MIC screening breakpoints (Table 1, Figures S2b, S2d). The three laboratories with the lowest performance observed elevated carbapenem MICs by AST initially. Subsequent ‘normal’ Etest carbapenem MICs led to the termination of further confirmation. In each of these three labs, detection failure of at least one CPE could be attributed to Etest carbapenem MICs below the screening breakpoint, despite previously increased carbapenem MICs by AST (Tables S2a, S2c, S2e, Figure S2a, S2c, S2e).

Discussion

Laboratories that used a CPE screening meropenem MIC breakpoint of >0.25 mg/L detected no or fewer OXA-244-producing E. coli compared to laboratories that followed their own testing algorithm. This screening breakpoint contributes to detection failures: OXA-244-producing E. coli isolates frequently have meropenem MICs <0.25 mg/L. Implementing the one-step lower EUCAST CPE screening recommendation (meropenem MIC >0.125 mg/L) improves CPE detection. However, an AST card used by three of the five labs from this study does not accommodate CPE detection with a meropenem MIC of 0.25 mg/L because that MIC is reported as ≤0.25 mg/L. Recent European emergence of bla_OXA-244, and the potential of OXA-244-producing strains with initial low carbapenem MICs to develop elevated MICs during carbapenem therapy, may warrant renewed optimization of CPE screening sensitivity versus specificity.^10,13,14

Weaknesses of this study include the low number of isolates and participating labs, and the use of OXA-244-producing E. coli isolates that had previously not been detected as CPE by one laboratory, i.e., were difficult to detect and may not be a representative sample of all OXA-244-producing E. coli. Neither the specificity of screening protocols nor non-E. coli CPE were studied. A strength of this study is the detailed disclosure of screening protocols by participating labs, thereby demonstrating performance variability across protocols.

For laboratories that rely on AST, to optimally detect OXA-244-producing E. coli, we recommend the advised EUCAST meropenem screening breakpoint, i.e. confirmatory testing in Enterobacterales with meropenem MIC >0.125 mg/L. AST cards should be able to detect meropenem MICs as low as 0.25 mg/L. Based on our data, but dependent on additional screening criteria, confirmatory carbapenem Etest will lead to unwanted failures of OXA-244-producing E. coli detections. Shortly before study participation, laboratory 4 had started to include a piperacillin/tazobactam MIC criterion for CPE confirmatory testing and was now among the top two performing laboratories in detection rates of OXA-244-producing E. coli in this study. OXA-244 efficiently hydrolyses piperacillin but should be partially susceptible to tazobactam inhibition.⁴ Nevertheless, piperacillin/tazobactam resistance is common in OXA-244-producing E. coli.^22–24 Although a piperacillin/tazobactam resistance screening criterium was clearly sensitive for OXA-244-producing E. coli detection in this study and is also part of a EUCAST-recommended algorithm for CPE, the cost-effectiveness of adding this criterium was not evaluated in this study, and the specificity of this criterium is likely low.²⁵ Although suggested by EUCAST, temocillin resistance was not part of the CPE screening algorithms in the participating laboratories. Temocillin MIC’s in OXA-244-producing E. coli are lower than other OXA-48-producing E. coli isolates but remain in the resistant MIC range.^25,26 Therefore, adding a temocillin resistance screening criterium might be beneficial. The CPO detect screening test, included in some commercial AST panels, performed well for this subset of OXA-244-producing E. coli isolates, in line with a recent report.²⁷ To treat and control CPE, it is of utmost importance to have reliable detection methods. Rise in difficult-to-detect OXA-244 carbapenemase may be an impetus for laboratories to modify their testing algorithms. Adherence to current guidelines does not guarantee the most sensitive testing algorithms.

Funding

This study was carried out as part of our routine work.

Transparency declarations

None to declare.

Declaration of generative AI and AI-assisted technologies in the writing process

Statement: During the preparation of this work, the author(s) used ChatGPT in order to improve readability and language. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

Supplementary data

Figures S1 to S2 and Tables S1 to S2 are available as Supplementary data at JAC-AMR Online.