This study aimed to determine whether agar dilution, research-use-only disk diffusion (Mast Group Ltd., Bootle Merseyside, UK), Etest (bioMérieux, Inc., Durham, NC), and MIC test strip (MTS) (Liofilchem, Inc., Waltham, MA) methods yield equivalent results to those of broth microdilution (BMD) for imipenem-relebactam susceptibility testing using a collection of 297 Gram-negative bacilli, including members of the order Enterobacterales and Pseudomonas aeruginosa, enriched for drug resistance.
KEYWORDS: imipenem-relebactam, broth microdilution, agar dilution, gradient strips, disk diffusion, susceptibility testing
ABSTRACT
This study aimed to determine whether agar dilution, research-use-only disk diffusion (Mast Group Ltd., Bootle Merseyside, UK), Etest (bioMérieux, Inc., Durham, NC), and MIC test strip (MTS) (Liofilchem, Inc., Waltham, MA) methods yield equivalent results to those of broth microdilution (BMD) for imipenem-relebactam susceptibility testing using a collection of 297 Gram-negative bacilli, including members of the order Enterobacterales and Pseudomonas aeruginosa, enriched for drug resistance. MIC and disk diameter results were interpreted using United States Food and Drug Administration breakpoints. Overall, 76.8% of the isolates tested were susceptible to imipenem-relebactam by BMD. MIC values for agar dilution, Etest, and MTS were not significantly different from that for BMD, although they tended to be 1 to 2 dilutions higher. Essential agreement was 95.6% for agar dilution, 90.6% for Etest, and 85.2% for MTS. Categorical agreement was 98.0% for agar dilution, 73.1% for disk diffusion, 96.3% for Etest, and 96.6% for MTS. In conclusion, agar dilution and Etest yielded comparable results to BMD for imipenem-relebactam.
INTRODUCTION
Infections caused by Gram-negative bacilli (GNB) are a rising threat because the prevalence of drug-resistant isolates is increasing over time (1–5). Drug-resistant GNB are encountered across many settings, including in intensive care and neonatal units (6) as well as in the community (7). Patients with infections caused by drug-resistant GNB may have higher mortality rates and longer hospital stays than those with infections by more susceptible isolates (8); these infections can be challenging to treat. New antibiotics are needed.
Imipenem was the first carbapenem introduced into clinical practice, but since its introduction, resistance has emerged (9). Relebactam is a β-lactamase inhibitor coformulated with imipenem as imipenem-relebactam to restore imipenem’s activity in the presence of Ambler class A and C β-lactamases (10–13). The combination has been approved by the United States Food and Drug Administration (FDA) for complicated urinary tract infections, including pyelonephritis, and for complicated intra-abdominal infections in adults.
This study aimed to evaluate the performance of agar dilution, research-use-only disk diffusion (Mast Group Ltd., Bootle Merseyside, UK), Etest (bioMérieux, Inc., Durham, NC), and MIC test strip (MTS) (Liofilchem, Inc., Waltham, MA) methods compared to that of broth microdilution (BMD) for imipenem-relebactam susceptibility testing using a collection of 297 drug-resistant GNB, including members of the order Enterobacterales, alongside Pseudomonas aeruginosa.
MATERIALS AND METHODS
Bacterial isolates.
A total of 297 isolates of drug-resistant GNB collected from the United States, Canada, and Singapore were studied. Isolates were previously characterized with regard to genetic resistance mechanisms by PCR, with or without sequencing, and were found to have the following resistance genes (number of isolates): blaKPC (106), blaCTX-M (58), blaCMY (21), blaIMP (11), blaNDM (31), blaOXA (18), blaSHV (28), blaTEM (40), blaIMI (5), blaVIM (3), blaSME (3), and blaFOX (1), with some harboring more than one resistance mechanism. The isolates studied represented Klebsiella pneumoniae complex (147), Escherichia coli (89), Enterobacter cloacae complex (21), Serratia marcescens (5), Citrobacter freundii (4), Citrobacter koseri (3), Klebsiella aerogenes (2), Citrobacter sedlakii, and P. aeruginosa (25).
Isolates had been stored in MicroBank vials (Pro-Lab Diagnostics, Round Rock, TX) at −80°C. Bacteria were grown on BBL Trypticase soy agar with 5% sheep blood (Becton, Dickinson, Franklin Lakes, NJ) and subcultured a second time to yield an F2 generation, which was tested. Antibiotics were not used to selectively maintain resistance. Quality control (QC) strains E. coli ATCC 25922, P. aeruginosa ATCC 27853, and K. pneumoniae ATCC 700603 were included in each test run. All results for QC strains were within acceptable Clinical and Laboratory Standards Institute (CLSI) ranges for QC isolates (14). Agar dilution and broth microdilution were performed using the same inoculum. Gradient strips and disk diffusion methods were conducted using the same inoculum. If repeated testing was needed for any of the methods compared to broth microdilution, these were conducted using the same inoculum.
Antimicrobial agent preparation.
Stock solutions at concentrations of 1,000 μg/ml in phosphate buffer at pH of 7.2 were prepared from imipenem and relebactam powders on a weekly basis.
Agar dilution.
Mueller-Hinton agar (MHA) plates were prepared according to CLSI recommendations (15). Final concentrations of imipenem ranged from 0.03 to 32 μg/ml in doubling dilution intervals with a constant concentration of relebactam of 4 μg/ml. Two antimicrobial-free MHA growth control plates were included in each run. Plates were prepared at the beginning of each week, stored at 4°C, and used within 3 days. A 0.5 McFarland bacterium-saline solution was made, of which 100 μl was mixed with 900 μl of cation-adjusted Mueller-Hinton broth (CAMHB) to yield a 1:10 dilution. From this dilution, 1 μl was used to manually inoculate plates, resulting in a bacterial density of 104 CFU/spot. Plates were then placed in a room air incubator at 37°C for 16 to 20 h and read according to CLSI recommendations (15).
Broth microdilution.
Broth microdilution was performed using the same final imipenem-relebactam concentrations used for agar dilution. One well per isolate was an antimicrobial-free growth control well. Broth microdilution plates were freshly prepared on the day of use. Broth microdilution and agar dilution were performed on the same day for individual isolates, using the same 0.5 McFarland bacterium-saline solution diluted to a 1:100 concentration by mixing 20 μl of bacterium-saline solution with 1,980 μl of CAMHB. Fifty microliters of the 1:100 diluted inoculum was added to each well, resulting in a bacterial density of 5 × 104 CFU/well. Plates were placed in a room air incubator at 37°C for 16 to 20 h and read according to CLSI recommendations (15). Purity checks of inocula were performed by incubating purity plates at 37°C and verifying the presence of pure colonies after 20 to 24 h of incubation.
Disk diffusion.
Disk diffusion testing was performed using research-use-only imipenem-relebactam disks (Mast Group Ltd., Bootle Merseyside, UK) (disk potency: imipenem, 10 μg; relebactam, 25 μg) according to the manufacturer’s procedures and CLSI methodology, with Mueller-Hinton II agar (Becton, Dickinson, Franklin Lakes, NJ). Plates were inoculated with samples of each isolate adjusted to a turbidity of 0.5 McFarland. Disks were dispensed onto the surface of the inoculated agar, and plates incubated for 16 to 18 h at 37°C. After incubation, diameters of zones of inhibition were measured using a digital caliper (Traceable; Fisher Scientific, Pittsburgh, PA), recorded, and interpreted according to FDA-identified breakpoints.
Gradient strips.
Two research-use-only gradient strips were studied: Etest strips (bioMérieux, Inc., Durham, NC) (strip potency for imipenem-relebactam of 0.002 to 32/4 μg/ml), stored at 4 to 8°C until use, and MTS (Liofilchem, Inc., Waltham, MA) (strip potency for imipenem-relebactam of 0.002 to 32/4 μg/ml), stored at −20°C until use. MICs were read after incubation, according to the instructions of the manufacturers, where the elliptical zone of inhibition intersects with the MIC scale on the strip. Recorded MICs were rounded up to the next 2-fold dilution, as recommended by the manufacturers. MIC breakpoints determined by the FDA were used for interpreting the collected data.
Data analysis.
The methods tested were interpreted using established FDA breakpoints for Enterobacterales as follows: a MIC of ≤1 μg/ml was interpreted as susceptible, a MIC of 2 μg/ml as intermediate, and a MIC of ≥4 μg/ml as resistant. Breakpoints used for P. aeruginosa were as follows: a MIC of ≤2 μg/ml was interpreted as susceptible, a MIC of 4 μg/ml as intermediate, and a MIC of ≥8 μg/ml as resistant. For disk diffusion, zone size interpretation according to FDA was as follows: ≥25 mm, susceptible; 21 to 24 mm, intermediate; and ≤20 mm, resistant for Enterobacterales; and ≥23 mm, susceptible; 20 to 22 mm, intermediate; and ≤19 mm, resistant for P. aeruginosa.
Essential agreement was assessed by calculating the percentage of isolates tested by the agar dilution and gradient strips that produced MICs within ±1 doubling dilutions compared to that with the reference method BMD. Essential agreement of ≥90% was considered acceptable (16). Categorical agreement was assessed by calculating the percentage of isolates tested by agar dilution, disk diffusion, and gradient strips that yielded the same susceptibility category as the reference method BMD. A percentage of ≥90% was considered acceptable. Minor errors (mEs), major errors (MEs), and very major errors (VMEs) were further assessed. Acceptable percentages of errors were ≤1.5% for VMEs, ≤3% for MEs, and ≤7% for mEs and MEs combined (17, 18).
Reproducibility testing was performed using four QC strains (E. coli ATCC 25922, P. aeruginosa ATCC 27853, E. coli ATCC 35218, and K. pneumoniae ATCC 700603) and two resistant K. pneumoniae complex isolates (both containing the NDM resistance mechanism) that were part of the study. Three inocula were produced daily of these six organisms, yielding 90 reproducibility testing results for each method. Acceptable reproducibility results were >95% for precision essential agreement (PEA) and precision categorical agreement (PCA), with ≤1 out-of-range result per QC strain.
Statistical analysis of MIC distributions determined by agar dilution and gradient strips versus BMD was performed using the signed-rank Wilcoxon test to assess trends in MIC disagreement. A P value of <0.05 was considered statistically significant. The Fisher exact test was performed to evaluate associations between errors and specific bacterial species. Analyses of agreement between BMD and all four methods tested were evaluated with Cohen’s kappa statistics graded from poor to perfect agreement (19). Additionally, an intraclass correlation coefficient was used to determine the correlation between the four susceptibility methods and BMD and interpreted according to Koo and Li (20).
Repeat testing.
Repeat testing was performed if there was no growth in the BMD growth control well. All testing associated with errors, including VMEs, MEs, and mEs, was repeated by BMD and the test method producing the error. Testing yielding results that were not consistent with expected results for imipenem-relebactam according to β-lactamase type (e.g., blaKPC-positive GNB testing resistant to imipenem-relebactam or blaIMP-positive GNB testing susceptible to imipenem-relebactam) (21) was repeated. Results are reported as the results of initial testing or, if repeat testing was performed, those of repeat testing.
RESULTS
Two-hundred ninety-seven isolates of drug-resistant GNB were tested for susceptibility to imipenem-relebactam by BMD, agar dilution, disk diffusion, Etest, and MTS (see Tables S1 and S2 in the supplemental material). The proportions of isolates categorized as resistant, intermediate, and susceptible to imipenem-relebactam according to FDA breakpoints (17) by each testing method are provided in Table 1. Performance of agar dilution, disk diffusion, Etest, and MTS in comparison to BMD is summarized in Table 2.
TABLE 1.
Interpretive results of imipenem-relebactam susceptibility by broth microdilution, agar dilution, disk diffusion, Etest, and MIC test strips for 297 isolates of Gram-negative bacilli
| Testing methoda | No. (%) of isolates with a result of: |
||
|---|---|---|---|
| Susceptible | Intermediate | Resistant | |
| Enterobacterales (n = 272) | |||
| BMDb | 208 (76.5) | 11 (4.0) | 53 (19.5) |
| Agar dilution | 212 (77.9) | 7 (2.6) | 53 (19.5) |
| Disk diffusionc | 141 (51.8) | 66 (24.3) | 65 (23.9) |
| Etest | 212 (77.9) | 10 (3.7) | 50 (18.4) |
| MTS | 211 (77.6) | 14 (5.1) | 47 (17.3) |
| P. aeruginosa (n = 25) | |||
| BMDd | 21 (84.0) | 0 (0) | 4 (16.0) |
| Agar dilution | 20 (80.0) | 1 (4.0) | 4 (16.0) |
| Disk diffusione | 21 (84.0) | 0 (0) | 4 (16.0) |
| Etest | 21 (84.0) | 0 (0) | 4 (16.0) |
| MTS | 21 (84.0) | 0 (0) | 4 (16.0) |
BMD, broth microdilution; MTS, MIC test strips.
FDA breakpoints for Enterobacterales by BMD: ≤1 μg/ml, susceptible; 2 μg/ml, intermediate; ≥4 μg/ml, resistant.
FDA breakpoints for Enterobacterales by disk diffusion: ≥25 mm, susceptible; 21 to 24 mm, intermediate; ≤20 mm, resistant.
FDA breakpoints for P. aeruginosa by BMD: ≤2 μg/ml, susceptible; 4 μg/ml, intermediate; ≥8 μg/ml, resistant.
FDA breakpoints for P. aeruginosa by disk diffusion: ≥23 mm, susceptible; 20 to 22 mm, intermediate; and ≤19 mm, resistant.
TABLE 2.
Performance of agar dilution, disk diffusion, Etest, and MIC test strips compared to broth microdilution for 297 isolates of Gram-negative bacilli
| Testing method | % agreement |
No. (%) of isolates with:a
|
|||
|---|---|---|---|---|---|
| Essential | Categorical | VME | ME | mE | |
| Agar dilution | 95.6 | 98.0 | 0 (0) | 0 (0) | 7 (2.4) |
| Disk diffusion | NAb | 74.1 | 0 (0) | 6 (2.0) | 73 (24.6) |
| Etest | 90.0 | 96.2 | 0 (0) | 0 (0) | 11 (3.7) |
| MTSc | 85.2 | 96.6 | 0 (0) | 0 (0) | 10 (3.4) |
VME, very major error; ME, major error; mE, minor error.
NA, not applicable.
MTS, MIC test strip.
Essential agreement.
Compared to BMD, agar dilution and Etest met the essential agreement (EA) cutoff rate of ≥90%, at 95.6% and 90.0%, respectively, whereas MTS fell short of being acceptable, at 85.2%. Compared to that of BMD, agar dilution produced higher MICs for 72 isolates (24.2%) and lower MICs for 57 (19.2%) (Fig. 1a), Etest produced higher MICs for 84 isolates (28.3%) and lower MICs for 62 (20.9%) (Fig. 1b), and MTS produced higher MICs for 96 isolates (32.3%) and lower MICs for 64 (21.5%) (Fig. 1c). A Wilcoxon signed-rank test performed on the distributions of MICs determined by agar dilution, Etest, and MTS versus BMD showed that the distributions of imipenem-relebactam MICs by all three methods were not higher than that by BMD (one-tailed P values: 0.696 for agar dilution, 0.505 for Etest, and 0.936 for MTS). Statistical analysis with an intraclass correlation coefficient showed almost perfect correlation between BMD and the study methods, with intraclass correlation coefficient (2k) values of 0.98, 0.95, and 0.93 for agar dilution, Etest, and MTS, respectively.
FIG 1.
MICs determined by broth microdilution and agar dilution (a), Etest (b), and MIC test strips (MTS) (c). White, congruent results; light gray, results within ±1 double dilution; dark gray, results more than ±1 double dilution; S, susceptible; R, resistant; I, intermediate.
Categorical agreement.
Compared to BMD, agar dilution, Etest, and MTS methods met the categorical agreement (CA) cutoff of ≥90% before and after repeat testing, with CA being 98.0%, 96.2%, and 96.6% for agar dilution, Etest, and MTS, respectively. Disk diffusion did not meet the cutoff criterion, with a CA of 74.1%. Error rates for the tested methods are shown in Table 2. Agar dilution, Etest, and MTS methods reached the cutoff criteria for very major error (VME) of ≤1.5%, major error (ME) of ≤3%, and minor error (mE) and ME combined of ≤7% after repeat testing, whereas with disk diffusion, 73 mEs (24.6%) and 6 MEs (2.0%) occurred. No VMEs were noted. mE rates were 2.4%, 3.7%, and 3.4% for agar dilution, Etest, and MTS, respectively. The agreement values of all four methods compared to BMD for CA Cohen’s kappa coefficients were 0.93, 0.89, 0.91, and 0.52 for agar dilution, Etest, MTS, and disk diffusion, respectively, with values closest to 1 considered near perfect agreement.
Essential and categorical agreement by species/complex.
Essential and categorical agreement by species/complexes is shown in Table S4. EA and CA were ≥90% for all methods for Escherichia coli, of which there were 89 isolates studied. Among the 147 Klebsiella pneumoniae complex isolates studied, EA was 95.2%, 87.8% and 83.0% for agar dilution, Etest, and MTS, respectively, and CA was 99.3%, 59.9%, 95.9%, and 97.3% for agar dilution, disk diffusion, Etest, and MTS, respectively.
Essential agreement and categorical agreement by resistance mechanism.
Essential agreement was ≥90% for isolates harboring blaCTX-M (53; 16 of which were also positive for additional non-extended-spectrum β-lactamase [ESBL] genes blaTEM-1, blaSHV-1, blaSHV-11, and blaOXA-1) for agar dilution, Etest, and MTS (Table 3). For isolates harboring blaKPC (106), EA was ≥90% for agar dilution and Etest. Performance in detecting susceptibility in blaKPC-positive isolates was poorer with the MTS method, with EA of 84.9%, with 90 of 106 MICs being within ±1 doubling dilution. For isolates positive for blaNDM (31), EA was ≥90% for agar dilution only, being 74.2% and 70.9% for Etest and MTS, respectively.
TABLE 3.
Essential and categorical agreement and error types for isolates harboring blaCTX-M, blaKPC, and blaNDM for agar dilution, disk diffusion, Etest, and MIC test strips compared to broth microdilution
| Isolatea | % agreement |
Error type (n [%])b
|
|||
|---|---|---|---|---|---|
| Essential | Categorical | VME | ME | mE | |
| Agar dilution | |||||
| blaCTX-M (n = 53) | 100 | 100 | 0 (0) | 0 (0) | 0 (0) |
| blaKPC (n = 106) | 97.2 | 99.1 | 0 (0) | 0 (0) | 1 (1) |
| blaNDM (n = 31) | 100 | 100 | 0 (0) | 0 (0) | 0 (0) |
| Etest | |||||
| blaCTX-M (n = 53) | 94.3 | 98.1 | 0 (0) | 0 (0) | 1 (1.9) |
| blaKPC (n = 106) | 94.3 | 98.1 | 0 (0) | 0 (0) | 2 (1.9) |
| blaNDM (n = 31) | 74.2 | 100 | 0 (0) | 0 (0) | 0 (0) |
| MTS | |||||
| blaCTX-M (n = 53) | 96.2 | 98.1 | 0 (0) | 0 (0) | 1 (1.9) |
| blaKPC (n = 106) | 84.9 | 97.2 | 0 (0) | 0 (0) | 3 (2.8) |
| blaNDM (n = 31) | 70.9 | 100 | 0 (0) | 0 (0) | 0 (0) |
| Disk | |||||
| blaCTX-M (n = 53) | NAc | 100 | 0 (0) | 0 (0) | 0 (0) |
| blaKPC (n = 106) | NA | 45.3 | 0 (0) | 2 (1.9) | 56 (52.8) |
| blaNDM (n = 31) | NA | 100 | 0 (0) | 0 (0) | 0 (0) |
blaCTX-M, cefotaxime-hydrolyzing β-lactamase (including 16 with additional non-ESBL genes blaTEM-1, blaSHV-1, blaSHV-11, and blaOXA-1); blaKPC, Klebsiella pneumoniae carbapenemase gene; blaNDM, New Delhi metallo-β-lactamase gene.
VME, very major error; ME, major error; mE, minor error.
NA, not applicable.
Categorical agreement was ≥90% for the blaCTX-M isolates for agar dilution, Etest, MTS, and disk diffusion. For isolates positive for blaKPC, CA was ≥90% for agar dilution, Etest, and MTS. Of the mE for blaKPC isolates, 56 were observed with disk diffusion, resulting in a CA of 45.3%. The CA for isolates positive for blaNDM was 100% for all four methods tested against BMD.
For the other resistance mechanisms tested, there were fewer than 30 isolates per grouping; therefore, their results are not separately reported. However, of interest, the blaOXA-positive isolates included 4 narrow-spectrum β-lactamases and 14 carbapenemases. In comparing Etest to BMD, there were 2 mEs, one each from a blaOXA-232- and blaOXA-48-positive isolate. In comparing disk diffusion to BMD, there were 4 MEs (3 from blaOXA-48-positive isolates and 1 from a blaOXA-181-positive isolates) and 5 mEs (4 from blaOXA-48-isolates and 1 from a blaOXA-232-positive isolate).
Reproducibility.
For reproducibility testing (Table 4), the precision essential agreement (PEA) values for agar dilution, Etest, and MTS compared to BMD were 100%, 98.9%, and 53.3%, respectively. The precision categorical agreement (PCA) for agar dilution, Etest, MTS, and disk diffusion compared to BMD was 100% for all. The MIC and zone size ranges for each QC strain and testing method are shown in Table 5. The MTS MIC ranges for E. coli ATCC 25922, E. coli ATCC 35218, K. pneumoniae ATCC 700603, and P. aeruginosa were consistently higher than the BMD MIC range by 1 to 3 doubling dilutions. Of the 90 test results for MTS, MICs trended higher than BMD, with 26 replicates +1, 40 replicates +2, and 2 replicates +3 dilutions from BMD. Etests also trended higher than BMD, with 47 replicates +1 and 1 replicate +2 dilutions from BMD. Agar dilution had a similar trend, with 26 replicates +1 dilution from BMD.
TABLE 4.
Essential and categorical agreement for reproducibility testing for agar dilution, disk diffusion, Etest, and MIC test strips compared to broth microdilution
| Testing method | % agreement |
|
|---|---|---|
| Essential | Categorical | |
| Agar dilution | 100 | 100 |
| Disk diffusion | NAa | 100 |
| Etest | 98.9 | 100 |
| MTSb | 53.3 | 100 |
NA, not applicable.
MTS, MIC test strip.
TABLE 5.
MIC and zone size ranges for each quality control strain and testing method
| Testing methoda | Value or range (μg/ml or mm) |
|||||
|---|---|---|---|---|---|---|
|
Escherichia coli |
Pseudomonas aeruginosa |
Klebsiella pneumoniae |
||||
| ATCC 25922 | ATCC 35218 | ATCC 27853 | ATCC 700603 | IDRL-10408 (blaNDM) | IDRL-10486 (blaNDM) | |
| BMD | 0.125–0.25 | 0.125 | 0.25–0.5 | 0.125–0.25 | >32 | 32 to >32 |
| Agar dilution | 0.125–0.25 | 0.125–0.25 | 0.5 | 0.125–0.25 | 32 to >32 | 32 to >32 |
| Disk diffusion | 25.2–30.08 | 25.99–30.27 | 26.82–29.85 | 24.81–29.26 | 0 | 10.27–13.63 |
| Etest | 0.125–0.25 | 0.125–0.25 | 0.5–1 | 0.125–0.25 | >32 | >32 |
| MTS | 0.5 | 0.5–1 | 1–2 | 0.25–0.5 | >32 | >32 |
BMD, broth microdilution; MTS, MIC test strip.
DISCUSSION
Infections with GNB have become an increasing challenge in clinical practice due to the emergence of resistance to many older antibiotics (22). Imipenem-relebactam is a new antibiotic with in vitro activity against several drug-resistant GNB. It is essential that laboratories have methods to reliably test susceptibility to this new agent. This study shows that agar dilution, alongside Etest strips, yield comparable results to those of BMD for imipenem-relebactam susceptibility testing of Enterobacterales and P. aeruginosa. Although BMD is considered the gold standard for testing antimicrobial susceptibility, having alternative methods available is a benefit to clinical laboratories and therefore to patients.
Overall, disk diffusion produced the highest number of mEs and had the lowest CA of the methods studied, with 73 mEs (24.6%), 6 MEs (2.0%), and a CA of 74.1%. Of the 73 mEs, 56 were associated with blaKPC-positive isolates. Sixty-two of 73 mEs were the result of disk diffusion categorization as intermediate and BMD categorization as susceptible (see Fig. S1 in the supplemental material). Based on these findings, it may be prudent to consider MIC testing for isolates testing intermediate by disk diffusion, similar to the recent recommendation for testing ceftazidime-avibactam (23). Determining resistance and susceptibility for imipenem-relebactam with disks and both strips for K. pneumoniae isolates was a challenge in some cases, because colonies grew within the zones of inhibition; the MIC was measured to the higher value or zone to the lower value. This has been described in other studies testing carbapenem resistance in K. pneumoniae carbapenemase (KPC)-producing Enterobacter isolates (21).
Categorizing susceptible isolates as intermediate may be clinically problematic, as clinicians may forego use of an antibiotic which could be useful. The intermediate category can be interpreted as activity of a drug with uncertain therapeutic effects and may reflect an isolate that may be appropriately treated if present in body sites where the drug is physically concentrated or that may be treated when a high dosage of drug can be used (14). Additionally, it can be considered a buffer zone that should prevent technical factors from causing major discrepancies in interpretations (24).
The performance of phenotypic susceptibility testing methods such as MTS and disk diffusion depends on a number of factors, such as storage, choice of material, and inoculation method, and are therefore prone to technical and human errors (25). Automation of antimicrobial susceptibility methods might improve accuracy and reduce error rates (26).
As for the reproducibility testing, the EA for MTS compared to BMD was quite low (53%) and did not meet the cutoff that 95% of results should be within EA. According to CLSI, some isolates, particularly those that produce β-lactamases, may produce MICs that vary over four or more dilutions for reference broth dilution testing (16), which could be the case here. CA for the reproducibility testing comparing all four methods to BMD yielded 100% agreement.
Several in vitro studies testing the activity of imipenem-relebactam against GNB have shown activity against GNB with K. pneumoniae carbapenemase, combinations of β-lactamases, and impermeability and P. aeruginosa expressing OprD mutants (10, 27–29). In contrast, relebactam does not potentiate imipenem’s activity against isolates harboring Ambler class B β-lactamases, which include New Delhi metallo-β-lactamase (NDM) and imipenemase metallo-β-lactamase (IMP) (30). This trend is reflected by our study, with blaKPC-positive isolates having susceptibility ranging from 87.7% to 94.3% across methods. Of note is that six blaKPC-positive isolates in our collection (4 K. pneumoniae complex, 1 S. marcescens, and 1 P. aeruginosa isolate) were unexpectedly not susceptible to imipenem-relebactam by all methods tested. An explanation for this unusual finding could be that these isolates harbor additional β-lactamases (e.g., IMP, Verona integron-encoded metallo-β-lactamase [VIM], or NDM) or efflux pumps and/or have porin mutations.
In our study, one isolate was IMP positive yet imipenem-relebactam susceptible. This could have been because of plasmid loss, as previously observed with storage and absence of antimicrobial pressure (31). A limitation of our study is that we used a frozen selection of isolates, and only a small number of isolates of some species was tested. Therefore, further studies are needed for specific bacterial groupings. An additional limitation is the use of research-use-only disks and strips; results in this study may not reflect the versions of these that are now approved/cleared by the FDA.
To conclude, agar dilution as well as Etest produced comparable results to BMD and may, therefore, be considered acceptable methods for testing susceptibility to imipenem-relebactam, whereas the performance of MTS and disk diffusion needs further evaluation.
Supplementary Material
ACKNOWLEDGMENTS
We acknowledge Melissa J. Karau, Peggy C. Kohner, Nicolynn C. Cole, Scott A. Cunningham, and Matthew P. Murphy for technical advice. We thank the Mayo Clinic Clinical Bacteriology Lab, Donna J. Hata from the Mayo Clinic in Jacksonville, FL, James R. Johnson from the VA Medical Center in Minneapolis, MN, and Hennepin County Medical Center in Minneapolis, MN, Mary K. Hayden and Karen Lolans from Rush University Medical Center, and Paul C. Schreckenberger from Loyola, both in Chicago, IL, Patricia J. Simner from the Johns Hopkins University, George G. Zhanel and Daryl J. Hoban from the University of Manitoba, Partha Pratim De, Sanjay Ryan Menon, and Shawn Vasoo from Tan Tock Seng Hospital, and Koh Tse Hsien from Singapore General Hospital in Singapore for isolates included in this study.
Research reported in this publication was supported by Merck & Co.
R.P. reports grants from CD Diagnostics, BioFire, Curetis, Merck & Co., Contrafect, Hutchison Biofilm Medical Solutions, Accelerate Diagnostics, Allergan, EnBiotix, and The Medicines Company. R. Patel is or has been a consultant to Curetis, Specific Technologies, Selux Dx, GenMark Diagnostics, PathoQuest, and Qvella; monies are paid to Mayo Clinic. In addition, R.P. has a patent issued on Bordetella pertussis/parapertussis PCR, a patent on a device/method for sonication with royalties paid by Samsung to Mayo Clinic, and a patent issued on an antibiofilm substance. R.P. receives travel reimbursement and an editor’s stipend from ASM and IDSA and honoraria from the NBME, Up-to-Date and the Infectious Diseases Board Review Course. No conflicts of interest, financial or otherwise, are declared by the other authors.
H.H. and E.B. performed experiments, analyzed and interpreted results, and drafted the manuscript. J.N.M. performed a statistical analysis of the results. K.E.G.-Q., R.P., S.M.S.-M., and A.N.S. edited and revised the manuscript. R.P. conceived of and designed the research and approved the final version of the manuscript.
Footnotes
Supplemental material is available online only.
REFERENCES
- 1.Sader HS, Farrell DJ, Flamm RK, Jones RN. 2014. Antimicrobial susceptibility of Gram-negative organisms isolated from patients hospitalized in intensive care units in United States and European hospitals (2009–2011). Diagn Microbiol Infect Dis 78:443–448. doi: 10.1016/j.diagmicrobio.2013.11.025. [DOI] [PubMed] [Google Scholar]
- 2.Rodloff AC, Dowzicky MJ. 2017. Antimicrobial susceptibility among European Gram-negative and Gram-positive isolates collected as part of the tigecycline evaluation and surveillance trial (2004–2014). Chemotherapy 62:1–11. doi: 10.1159/000445022. [DOI] [PubMed] [Google Scholar]
- 3.Chang YT, Coombs G, Ling T, Balaji V, Rodrigues C, Mikamo H, Kim MJ, Rajasekaram DG, Mendoza M, Tan TY, Kiratisin P, Ni Y, Barry W, Xu Y, Chen YH, Hsueh PR. 2017. Epidemiology and trends in the antibiotic susceptibilities of Gram-negative bacilli isolated from patients with intra-abdominal infections in the Asia-Pacific region, 2010–2013. Int J Antimicrob Agents 49:734–739. doi: 10.1016/j.ijantimicag.2017.01.030. [DOI] [PubMed] [Google Scholar]
- 4.Veeraraghavan B, Jesudason MR, Prakasah JAJ, Anandan S, Sahni RD, Pragasam AK, Bakthavatchalam YD, Selvakumar RJ, Dhole TN, Rodrigues C, Roy I, Joshi S, Chaudhuri BN, Chitnis DS. 2018. Antimicrobial susceptibility profiles of Gram-negative bacteria causing infections collected across India during 2014–2016: study for monitoring antimicrobial resistance trend report. Indian J Med Microbiol 36:32–36. doi: 10.4103/ijmm.IJMM_17_415. [DOI] [PubMed] [Google Scholar]
- 5.Ponce-de-Leon A, Rodríguez-Noriega E, Morfín-Otero R, Cornejo-Juárez DP, Tinoco JC, Martínez-Gamboa A, Gaona-Tapia CJ, Guerrero-Almeida ML, Martin-Onraët A, Vallejo Cervantes JL, Sifuentes-Osornio J. 2018. Antimicrobial susceptibility of Gram-negative bacilli isolated from intra-abdominal and urinary-tract infections in Mexico from 2009 to 2015: results from the study for monitoring antimicrobial resistance trends (SMART). PLoS One 13:e0198621. doi: 10.1371/journal.pone.0198621. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Folgori L, Bielicki J, Heath PT, Sharland M. 2017. Antimicrobial-resistant Gram-negative infections in neonates: burden of disease and challenges in treatment. Curr Opin Infect Dis 30:281–288. doi: 10.1097/QCO.0000000000000371. [DOI] [PubMed] [Google Scholar]
- 7.Ironmonger D, Edeghere O, Bains A, Loy R, Woodford N, Hawkey PM. 2015. Surveillance of antibiotic susceptibility of urinary tract pathogens for a population of 5.6 million over 4 years. J Antimicrob Chemother 70:1744–1750. doi: 10.1093/jac/dkv043. [DOI] [PubMed] [Google Scholar]
- 8.Macgowan AP, BSAC Working Parties on Resistance Surveillance. 2008. Clinical implications of antimicrobial resistance for therapy. J Antimicrob Chemother 62 Suppl 2:ii105–ii114. doi: 10.1093/jac/dkn357. [DOI] [PubMed] [Google Scholar]
- 9.Remschmidt C, Schneider S, Meyer E, Schroeren-Boersch B, Gastmeier P, Schwab F. 2017. Surveillance of antibiotic use and resistance in intensive care units (SARI). Dtsch Arztebl Int 114:858–865. doi: 10.3238/arztebl.2017.0858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Schmidt-Malan SM, Mishra AJ, Mushtaq A, Brinkman CL, Patel R. 2018. In vitro activity of imipenem-relebactam and ceftolozane-tazobactam against resistant Gram-negative bacilli. Antimicrob Agents Chemother 62:e00533-18. doi: 10.1128/AAC.00533-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lob SH, Hackel MA, Kazmierczak KM, Hoban DJ, Young K, Motyl MR, Karlowsky JA, Sahm DF. 2017. In vitro activity of imipenem-relebactam against Gram-negative bacilli isolated from patients with lower respiratory tract infections in the United States in 2015 - results from the SMART global surveillance program. Diagn Microbiol Infect Dis 88:171–176. doi: 10.1016/j.diagmicrobio.2017.02.018. [DOI] [PubMed] [Google Scholar]
- 12.Karlowsky JA, Lob SH, Young K, Motyl MR, Sahm DF. 2018. Activity of imipenem/relebactam against Pseudomonas aeruginosa with antimicrobial-resistant phenotypes from seven global regions: SMART 2015–2016. J Glob Antimicrob Resist 15:140–147. doi: 10.1016/j.jgar.2018.07.012. [DOI] [PubMed] [Google Scholar]
- 13.Karlowsky JA, Lob SH, Kazmierczak KM, Hawser SP, Magnet S, Young K, Motyl MR, Sahm DF. 2018. In vitro activity of imipenem/relebactam against Gram-negative ESKAPE pathogens isolated in 17 European countries: 2015 SMART surveillance programme. J Antimicrob Chemother 73:1872–1879. doi: 10.1093/jac/dky107. [DOI] [PubMed] [Google Scholar]
- 14.CLSI. 2019. Performance standards for antimicrobial susceptibility testing, 29th ed CLSI standard M100 Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 15.CLSI. 2018. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. 11th ed CLSI standard M07 Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 16.CLSI. 2015. Verification of commercial microbial identification and antimicrobial susceptibility testing systems. M52 1st ed Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 17.FDA. 2009. Class II special controls guidance document: antimicrobial susceptibility test (AST) systems. US Food and Drug Administration Center for Devices and Radiological Health, Rockville, MD. [Google Scholar]
- 18.Clark RB, Lewinski MA, Loeffelholz MJ, Tibbetts RJ, Sharp SE. (ed) 2009. 31A. Verification and validation of procedures in the clinical microbiology laboratory (Cumitech). ASM Press, Washington, DC. [Google Scholar]
- 19.Viera AJ, Garrett JM. 2005. Understanding interobserver agreement: the kappa statistic. Fam Med 37:360–363. [PubMed] [Google Scholar]
- 20.Koo TK, Li MY. 2016. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 15:155–163. doi: 10.1016/j.jcm.2016.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Rechenchoski DZ, Dambrozio AML, Vivan ACP, Schuroff PA, Burgos TDN, Pelisson M, Perugini MRE, Vespero EC. 2017. Antimicrobial activity evaluation and comparison of methods of susceptibility for Klebsiella pneumoniae carbapenemase (KPC)-producing Enterobacter spp. isolates. Braz J Microbiol 48:509–514. doi: 10.1016/j.bjm.2017.01.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Peleg AY, Hooper DC. 2010. Hospital-acquired infections due to Gram-negative bacteria. N Engl J Med 362:1804–1813. doi: 10.1056/NEJMra0904124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sader HS, Rhomberg PR, Chandrasekaran S, Trejo M, Fedler KA, Boyken LD, Diekema DJ. 2020. Correlation between broth microdilution and disk diffusion results when testing ceftazidime-avibactam against a challenge collection of Enterobacterales isolates: results from a multilaboratory study. J Clin Microbiol 58:e01757-19. doi: 10.1128/JCM.01757-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hombach M, Böttger EC, Roos M. 2013. The critical influence of the intermediate category on interpretation errors in revised EUCAST and CLSI antimicrobial susceptibility testing guidelines. Clin Microbiol Infect 19:E59–E71. doi: 10.1111/1469-0691.12090. [DOI] [PubMed] [Google Scholar]
- 25.Hombach M, Ochoa C, Maurer FP, Pfiffner T, Böttger EC, Furrer R. 2016. Relative contribution of biological variation and technical variables to zone diameter variations of disc diffusion susceptibility testing. J Antimicrob Chemother 71:141–151. doi: 10.1093/jac/dkv309. [DOI] [PubMed] [Google Scholar]
- 26.Idelevich EA, Becker K. 2019. How to accelerate antimicrobial susceptibility testing. Clin Microbiol Infect 25:1347–1355. doi: 10.1016/j.cmi.2019.04.025. [DOI] [PubMed] [Google Scholar]
- 27.Livermore DM, Warner M, Mushtaq S. 2013. Activity of MK-7655 combined with imipenem against Enterobacteriaceae and Pseudomonas aeruginosa. J Antimicrob Chemother 68:2286–2290. doi: 10.1093/jac/dkt178. [DOI] [PubMed] [Google Scholar]
- 28.Toussaint KA, Gallagher JC. 2015. Beta-lactam/beta-lactamase inhibitor combinations: from then to now. Ann Pharmacother 49:86–98. doi: 10.1177/1060028014556652. [DOI] [PubMed] [Google Scholar]
- 29.Lee J-Y, Ko KS. 2012. OprD mutations and inactivation, expression of efflux pumps and AmpC, and metallo-β-lactamases in carbapenem-resistant Pseudomonas aeruginosa isolates from South Korea. Int J Antimicrob Agents 40:168–172. doi: 10.1016/j.ijantimicag.2012.04.004. [DOI] [PubMed] [Google Scholar]
- 30.Zhanel GG, Lawrence CK, Adam H, Schweizer F, Zelenitsky S, Zhanel M, Lagacé-Wiens PRS, Walkty A, Denisuik A, Golden A, Gin AS, Hoban DJ, Lynch JP, Karlowsky JA. 2018. Imipenem–relebactam and meropenem–vaborbactam: two novel carbapenem-β-lactamase inhibitor combinations. Drugs 78:65–98. doi: 10.1007/s40265-017-0851-9. [DOI] [PubMed] [Google Scholar]
- 31.Takahashi A, Yomoda S, Kobayashi I, Okubo T, Tsunoda M, Iyobe S. 2000. Detection of carbapenemase-producing Acinetobacter baumannii in a hospital. J Clin Microbiol 38:526–529. doi: 10.1128/JCM.38.2.526-529.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.