465
DIAGN MICROBIOLINFECTDIS 1992;15:465-468
The E-Test Challenged with Selected Strains Derek F.J. Brown
Results with the E-Test minimum inhibitory concentration (MIC) method correlated well with an agar dilution method in tests using a wide range of selected bacteria. Detection of methicillin resistance in Staphylococcus aureus was as relia-
ble as agar dilution and disk diffusion methods. The ease of use of the E-Test makes it particularly suitable for routine clinical microbiology laboratories.
INTRODUCTION
spp., 10 of Citrobacter spp., 20 of Enterobacter spp., 60 of Escherichia coli, 26 of Klebsiella spp., four of Morganella spp., 29 of Proteus spp., 12 of Serratia spp., 48 of Pseudomonas spp., nine of Xanthomonas maltophilia, 78 of S. aureus, 11 of S. epidermidis, 21 of enterococci, 39 of other streptococci, six of Neisseria gonorrhoeae, six of Bacillus spp., three of Clostridium spp., and three of Capnocytophaga spp. When possible, organisms were selected so that there was a distribution of susceptibility over the concentration range covered by the E-Test. Strains having MICs outside the E-Test range were excluded from the analysis. The 11 antimicrobial agents (number of strains tested) were amikacin (144), ampicillin (102), ciprofloxacin (251), cefotaxime (109), ceftazidime (141), erythromycin (131), gentamidn (241), penicillin (127), piperacillin (129), trimethoprim (109), and vancomycin (23). The agar dilution method used was similar to that described by the NCCLS (1990) except that Iso-Sensitest agar (Oxoid) was used, supplemented with 5% horse blood for tests on fastidious streptococci and 5% heated horse blood for tests on N. gonorrhoeae. For the E-Tests, the medium was the same as that used for the agar dilution method.
The E-Test is a novel method for measuring minimum inhibitory concentrations (MICs) of antimicrobial agents against bacteria (Bolmstr6m et al., 1988). Tests are set up by a similar technique to that used for disk diffusion tests, except that E-Test strips, which carry a preformed antibiotic gradient, are used instead of disks. Elliptical zones of inhibition are produced, and the MIC is read directly from the graduated E-Test strip at the point of intersection of the zone of inhibition with the strip. Up to six tests can be set up on a single 15-cm plate or single tests can be done on 9-cm plates. I report here comparison of the performance of the E-Test with an agar dilution MIC method, an examination of the E-Test for detection of methicillin resistance in Staphylococcus aureus, and experience with routine use of the E-Test.
MATERIALS A N D M E T H O D S For comparison of the E-Test with an agar dilution method, recent isolates and organisms from our culture collection were tested. The organisms included six strains of Acinetobacter spp., two of Aeromonas From the ClinicalMicrobiologyand Public Health Laboratory, AddenbrookeHospital, Cambridge, England, UK. Address reprint requests to Dr. D.F.J. Brown, ClinicalMicrobiology and Public Health Laboratory,AddenbrookeHospital, Hills Road, Cambridge CB2 2QW, UK. Received 3 December 1991; revised and accepted 20 December 1991. © 1992Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/92/$5.00
RESULTS A N D D I S C U S S I O N The correlation between the results of tests by agar dilution and E-Test methods for the 11 agents combined is shown in Figure 1 as the number of twofold dilution step differences between the E-Test and the agar dilution method results. The E-Test strip is graduated in half dilution steps, and we included half dilution step differences within the higher, full,
466
D.F.J. Brown
40 ¢~30
~20 m
OlO r
7-2
-2 -1 0 1 2 MIC difference (twofold dilutions)
i
i
~2
FIGURE 1 Correlation of MICs by E-Test and agar dilution in 1505 tests.
twofold dilution step differences in this distribution. Of the tests, 87.9% were within one log2 dilution, and tests differing by more than two log2 dilutions were rare (1%). With gentamicin, MICs tended to be higher by the E-Test, whereas, with ciprofloxacin, 60% of results were one loga dilution lower by the E-Test. Similar trends were noted by Baker et al., (1991). MICs of amikacin, ampicillin, cefotaxime, piperacillin, and trimethoprim were marginally lower by the E-Test. The correlation is slightly better than that reported previously with fewer test comparisons (Brown and Brown, 1991). Testing a different range of antimicrobial agents, Baker et al. (1991) reported 91.2% tests of Gram-positive organisms and 92.5% tests of Gram-negative organisms within one log2 dilution by the E-Test and agar dilution methods. Of the 15 results differing by more than fourfold, 13 were with penicillinase-producing strains of S. aureus tested against benzylpenicillin. Penicillin MICs by the agar dilution method are particularly inoculum dependent with penicillinase-producing staphylococci. The E-Test consistently gave higher penicillin MICs and was, therefore, considered a more reliable indicator of resistance. E-Test zones were generally clear and presented no problem in reading MIC end points. Unusual zones were seen with some penicillinase-producing staphylococci tested against penicillin and some erythromycin-resistant staphylococci tested against erythromycin, but MICs were reproducibly readable (Brown and Brown, 1991). Care must be taken to avoid overlooking small, resistant colonies close to where the zone intersects with the E-Test strip. This type of zone was particularly evident with two strains of X. maltophilia having ceftazidime MICs of 64 p,g/ml by agar dilution and small colonies up to 32 ~g/ml by E-Test (Figure 2). To a lesser extent, this effect was seen with a strain of Enterococcus faecium with a vancomycin MIC of 16 p,g/ml by agar dilution and small colonies up to 8 p~g/mlby the E-Test, and with
a strain of Streptococcus pneumoniae with a penicillin MIC of 0.5 p,g/ml by agar dilution and small colonies also up to 0.5 p~g/ml by the E-Test. Overall, the ETest correlated well with the agar dilution method for determination of MICs. It is, however, necessary to be aware that in a few situations with the E-Test small colonies within obvious zones of inhibition may be indicative of resistance. The E-Test was evaluated for detection of methicillin resistance in S. aureus by comparing the results of tests on 78 methicillin-resistant strains by E-Test, agar dilution, and disk diffusion methods. Tests were set up in parallel on Mueller-Hinton agar with and without 5% NaC1, with both methicillin and oxacillin, and with incubation at 30° and 37°C. The results will be published in detail elsewhere, but essentially none of the three methods detected methicillin-oxacillin resistance in all 78 strains under any combination of test conditions. In terms of detection of resistant strains, the E-Test was marginally more reliable than agar dilution in tests with oxacillin and vice versa with methicillin. The E-Test was easier to set up and read than were the agar dilution tests. The appearance of zones around E-Tests varied considerably and may relate to differences in expression of resistance seen in various types of methicillinoxacillin-resistant strains. Although sensitive strains had clear elliptical zones, resistance was seen as colonies within a zone of inhibition, gradually reducing colony size towards the E-Test strip, or double zones. The double zones were reminiscent of target zones occasionally seen with disk diffusion tests. Use of the E-Test in routine clinical laboratories will depend on how MICs are applied in particular laboratories. We have been using the E-Test routinely for 1 year and have found the method worthwhile for some isolates from blood cultures, for slowgrowing organisms, and to check unusual resistances or equivocal results in disk diffusion tests. In routine use in the United Kingdom, where single tests are commonly set up, a control strain can be included with each E-Test by inoculating one-half of a plate with the test organism and the other half of the plate with the control organism with a gap of 2 mm between the organisms. The E-Test strip is placed on the line between the two organisms (Figure 3). This approach is particularly useful as each E-Test strip is controlled, and there are financial savings as additional agar plates and E-Test strips are not used for control organisms. Inclusion of a control organism on the same plate is not appropriate in situations where resistance mechanisms are based on production of large amounts of diffusible enzymes, for example, hyperpenicillinase-producing staphylococci that may raise the penicillin MIC of the control strain on the same plate by two- to fourfold.
The E-Test C h a l l e n g e d with Selected Strains
FIGURE 2 Resistance of Xanthomonas maltophilia to ceftazidime demonstrated with the E-Test by small colonies at the intersection of the elliptical zone of inhibition and the E-Test strip.
467
468
D.F.J. B r o w n
FIGURE 3 Control of a ciprofloxacin E-Test by inoculating half of the plate with the control Escherichia NCTC 10418 (left) and half of the plate with the test organism, a strain of Citrobacter freundii (right). In conclusion, the E-Test yielded g o o d a g r e e m e n t with agar dilution m e t h o d s for determination of MICs. In the detection of methicillin resistance in S. aureus, the E-Test w a s as reliable as agar dilution a n d disk diffusion m e t h o d s a n d m a y p r o v i d e m o r e informa-
tion on expression of resistance t h a n either of the other m e t h o d s . The versatility a n d ease of u s e of the E-Test gives the m e t h o d considerable a p p e a l for use in routine clinical microbiology laboratories.
REFERENCES Baker CN, Stocker SA, Culver DH, Thornsberry C (1991) Comparison of the E test to agar dilution, broth microdilution, and agar diffusion susceptibility testing techniques by using a special challenge set of bacteria. J Clin Microbiol 29:533-538. Brown DFJ, Brown L (1991) Evaluation of the E test, a novel method of quantifying antimicrobial activity. J Antimicrob Chemother 27:185-190. Bolmstr6m A, Arvidson S, Ericsson M, Karlsson A (1988) A novel technique for direct quantification of antimi-
crobial susceptibility of microorganisms [abst 1209]. In Program and Abstracts of the 28th Interscience Conference on Antimicrobial Agents and Chemotherapy, Los Angeles, CA. Washington, DC: American Society for Microbiology. National Committee for Clinical Laboratory Standards (NCCLS) (1990) Approved Standard M7-A2: methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 2nd ed. Villanova, PA: NCCLS.
DIAGN MICROBIOLINFECTDIS 1992;15:465-468
The E-Test Challenged with Selected Strains Derek F.J. Brown
Results with the E-Test minimum inhibitory concentration (MIC) method correlated well with an agar dilution method in tests using a wide range of selected bacteria. Detection of methicillin resistance in Staphylococcus aureus was as relia-
ble as agar dilution and disk diffusion methods. The ease of use of the E-Test makes it particularly suitable for routine clinical microbiology laboratories.
INTRODUCTION
spp., 10 of Citrobacter spp., 20 of Enterobacter spp., 60 of Escherichia coli, 26 of Klebsiella spp., four of Morganella spp., 29 of Proteus spp., 12 of Serratia spp., 48 of Pseudomonas spp., nine of Xanthomonas maltophilia, 78 of S. aureus, 11 of S. epidermidis, 21 of enterococci, 39 of other streptococci, six of Neisseria gonorrhoeae, six of Bacillus spp., three of Clostridium spp., and three of Capnocytophaga spp. When possible, organisms were selected so that there was a distribution of susceptibility over the concentration range covered by the E-Test. Strains having MICs outside the E-Test range were excluded from the analysis. The 11 antimicrobial agents (number of strains tested) were amikacin (144), ampicillin (102), ciprofloxacin (251), cefotaxime (109), ceftazidime (141), erythromycin (131), gentamidn (241), penicillin (127), piperacillin (129), trimethoprim (109), and vancomycin (23). The agar dilution method used was similar to that described by the NCCLS (1990) except that Iso-Sensitest agar (Oxoid) was used, supplemented with 5% horse blood for tests on fastidious streptococci and 5% heated horse blood for tests on N. gonorrhoeae. For the E-Tests, the medium was the same as that used for the agar dilution method.
The E-Test is a novel method for measuring minimum inhibitory concentrations (MICs) of antimicrobial agents against bacteria (Bolmstr6m et al., 1988). Tests are set up by a similar technique to that used for disk diffusion tests, except that E-Test strips, which carry a preformed antibiotic gradient, are used instead of disks. Elliptical zones of inhibition are produced, and the MIC is read directly from the graduated E-Test strip at the point of intersection of the zone of inhibition with the strip. Up to six tests can be set up on a single 15-cm plate or single tests can be done on 9-cm plates. I report here comparison of the performance of the E-Test with an agar dilution MIC method, an examination of the E-Test for detection of methicillin resistance in Staphylococcus aureus, and experience with routine use of the E-Test.
MATERIALS A N D M E T H O D S For comparison of the E-Test with an agar dilution method, recent isolates and organisms from our culture collection were tested. The organisms included six strains of Acinetobacter spp., two of Aeromonas From the ClinicalMicrobiologyand Public Health Laboratory, AddenbrookeHospital, Cambridge, England, UK. Address reprint requests to Dr. D.F.J. Brown, ClinicalMicrobiology and Public Health Laboratory,AddenbrookeHospital, Hills Road, Cambridge CB2 2QW, UK. Received 3 December 1991; revised and accepted 20 December 1991. © 1992Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/92/$5.00
RESULTS A N D D I S C U S S I O N The correlation between the results of tests by agar dilution and E-Test methods for the 11 agents combined is shown in Figure 1 as the number of twofold dilution step differences between the E-Test and the agar dilution method results. The E-Test strip is graduated in half dilution steps, and we included half dilution step differences within the higher, full,
466
D.F.J. Brown
40 ¢~30
~20 m
OlO r
7-2
-2 -1 0 1 2 MIC difference (twofold dilutions)
i
i
~2
FIGURE 1 Correlation of MICs by E-Test and agar dilution in 1505 tests.
twofold dilution step differences in this distribution. Of the tests, 87.9% were within one log2 dilution, and tests differing by more than two log2 dilutions were rare (1%). With gentamicin, MICs tended to be higher by the E-Test, whereas, with ciprofloxacin, 60% of results were one loga dilution lower by the E-Test. Similar trends were noted by Baker et al., (1991). MICs of amikacin, ampicillin, cefotaxime, piperacillin, and trimethoprim were marginally lower by the E-Test. The correlation is slightly better than that reported previously with fewer test comparisons (Brown and Brown, 1991). Testing a different range of antimicrobial agents, Baker et al. (1991) reported 91.2% tests of Gram-positive organisms and 92.5% tests of Gram-negative organisms within one log2 dilution by the E-Test and agar dilution methods. Of the 15 results differing by more than fourfold, 13 were with penicillinase-producing strains of S. aureus tested against benzylpenicillin. Penicillin MICs by the agar dilution method are particularly inoculum dependent with penicillinase-producing staphylococci. The E-Test consistently gave higher penicillin MICs and was, therefore, considered a more reliable indicator of resistance. E-Test zones were generally clear and presented no problem in reading MIC end points. Unusual zones were seen with some penicillinase-producing staphylococci tested against penicillin and some erythromycin-resistant staphylococci tested against erythromycin, but MICs were reproducibly readable (Brown and Brown, 1991). Care must be taken to avoid overlooking small, resistant colonies close to where the zone intersects with the E-Test strip. This type of zone was particularly evident with two strains of X. maltophilia having ceftazidime MICs of 64 p,g/ml by agar dilution and small colonies up to 32 ~g/ml by E-Test (Figure 2). To a lesser extent, this effect was seen with a strain of Enterococcus faecium with a vancomycin MIC of 16 p,g/ml by agar dilution and small colonies up to 8 p~g/mlby the E-Test, and with
a strain of Streptococcus pneumoniae with a penicillin MIC of 0.5 p,g/ml by agar dilution and small colonies also up to 0.5 p~g/ml by the E-Test. Overall, the ETest correlated well with the agar dilution method for determination of MICs. It is, however, necessary to be aware that in a few situations with the E-Test small colonies within obvious zones of inhibition may be indicative of resistance. The E-Test was evaluated for detection of methicillin resistance in S. aureus by comparing the results of tests on 78 methicillin-resistant strains by E-Test, agar dilution, and disk diffusion methods. Tests were set up in parallel on Mueller-Hinton agar with and without 5% NaC1, with both methicillin and oxacillin, and with incubation at 30° and 37°C. The results will be published in detail elsewhere, but essentially none of the three methods detected methicillin-oxacillin resistance in all 78 strains under any combination of test conditions. In terms of detection of resistant strains, the E-Test was marginally more reliable than agar dilution in tests with oxacillin and vice versa with methicillin. The E-Test was easier to set up and read than were the agar dilution tests. The appearance of zones around E-Tests varied considerably and may relate to differences in expression of resistance seen in various types of methicillinoxacillin-resistant strains. Although sensitive strains had clear elliptical zones, resistance was seen as colonies within a zone of inhibition, gradually reducing colony size towards the E-Test strip, or double zones. The double zones were reminiscent of target zones occasionally seen with disk diffusion tests. Use of the E-Test in routine clinical laboratories will depend on how MICs are applied in particular laboratories. We have been using the E-Test routinely for 1 year and have found the method worthwhile for some isolates from blood cultures, for slowgrowing organisms, and to check unusual resistances or equivocal results in disk diffusion tests. In routine use in the United Kingdom, where single tests are commonly set up, a control strain can be included with each E-Test by inoculating one-half of a plate with the test organism and the other half of the plate with the control organism with a gap of 2 mm between the organisms. The E-Test strip is placed on the line between the two organisms (Figure 3). This approach is particularly useful as each E-Test strip is controlled, and there are financial savings as additional agar plates and E-Test strips are not used for control organisms. Inclusion of a control organism on the same plate is not appropriate in situations where resistance mechanisms are based on production of large amounts of diffusible enzymes, for example, hyperpenicillinase-producing staphylococci that may raise the penicillin MIC of the control strain on the same plate by two- to fourfold.
The E-Test C h a l l e n g e d with Selected Strains
FIGURE 2 Resistance of Xanthomonas maltophilia to ceftazidime demonstrated with the E-Test by small colonies at the intersection of the elliptical zone of inhibition and the E-Test strip.
467
468
D.F.J. B r o w n
FIGURE 3 Control of a ciprofloxacin E-Test by inoculating half of the plate with the control Escherichia NCTC 10418 (left) and half of the plate with the test organism, a strain of Citrobacter freundii (right). In conclusion, the E-Test yielded g o o d a g r e e m e n t with agar dilution m e t h o d s for determination of MICs. In the detection of methicillin resistance in S. aureus, the E-Test w a s as reliable as agar dilution a n d disk diffusion m e t h o d s a n d m a y p r o v i d e m o r e informa-
tion on expression of resistance t h a n either of the other m e t h o d s . The versatility a n d ease of u s e of the E-Test gives the m e t h o d considerable a p p e a l for use in routine clinical microbiology laboratories.
REFERENCES Baker CN, Stocker SA, Culver DH, Thornsberry C (1991) Comparison of the E test to agar dilution, broth microdilution, and agar diffusion susceptibility testing techniques by using a special challenge set of bacteria. J Clin Microbiol 29:533-538. Brown DFJ, Brown L (1991) Evaluation of the E test, a novel method of quantifying antimicrobial activity. J Antimicrob Chemother 27:185-190. Bolmstr6m A, Arvidson S, Ericsson M, Karlsson A (1988) A novel technique for direct quantification of antimi-
crobial susceptibility of microorganisms [abst 1209]. In Program and Abstracts of the 28th Interscience Conference on Antimicrobial Agents and Chemotherapy, Los Angeles, CA. Washington, DC: American Society for Microbiology. National Committee for Clinical Laboratory Standards (NCCLS) (1990) Approved Standard M7-A2: methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 2nd ed. Villanova, PA: NCCLS.
