Guidelines on susceptibility testing of antibiotic-resistant Enterobacteriaceae due to extended spectrum beta-lactamases (ESBLs)

Canadian External Quality Assessment
Advisory Group for Antibiotic Resistance

November 1998

ISBN 0-662-02429-X
© 1998

1.0 INTRODUCTION

These guidelines are produced under the auspices and authority of the Canadian External Quality Assessment - Advisory Group on Antibiotic Resistance (CEQA-AGAR). They represent a consensus of peer reviewed information and expert opinion on the most appropriate ways to test for and report antimicrobial susceptibility and resistance. Clinical situations and individual laboratory operations may necessitate some variations from these guidelines. These guidelines are strongly recommended for use by all laboratories providing information for national surveillance programs on antimicrobial resistance.

2.0 PREAMBLE

A common mechanism of bacterial resistance to beta-lactam antibiotics is the production of beta-lactamase enzymes that break down the structural beta-lactam ring of penicillin-like drugs. Genetic control of beta-lactamase production resides either on plasmids or on the chromosome, while expression is either constitutive or inducible.

Among the plasmid-encoded enzymes are TEM-1, TEM-2, and SHV-1 (Bush group 2b) that confer resistance to penicillins but not to the newer cephalosporins.⁽¹⁾,⁽²⁾ The first bacterial isolate resistant to the so-called extended spectrum cephalosporins was identified in Germany in 1983.⁽³⁾ Subsequently, much of the dramatic increase in bacterial resistance to beta-lactam antibiotics has been associated with the spread of extended spectrum beta-lactamases (ESBLs). These enzymes hydrolyze drugs such as ceftazidime, cefotaxime and aztreonam, but have little effect on the cephamycins - cefoxitin and cefotetan. ESBLs (group 2be) are encoded on mutated TEM-1 and SHV-1 genes carried on plasmids that are readily transmissible to other organisms.⁽⁴⁾ Relatively minor changes in the sequences of the original genes caused a significant alteration in the substrate affinity of the enzymes. Hydrolysis of ceftazidime, cefotaxime and aztreonam occurs at rates >10% of that for benzylpenicillin. These enzymes can be blocked by beta-lactamase inhibitors such as clavulanic acid.⁽⁵⁾

The prevalence of ESBL-producing bacteria in most hospitals remains unknown in spite of numerous reports of nosocomial outbreaks of infection due to these organisms. Difficulty in the detection of ESBL production using routine antimicrobial susceptibility testing methods has been well documented.⁽⁶⁾ Producing strains may appear to be susceptible.

3.0 DEFINITIONS

3.1 ESBL. Extended spectrum beta lactamase. This term refers to beta lactamase enzymes produced mainly by Klebsiella and E. coli that encode for resistance to broad-spectrum beta-lactam antibiotics that normally have activity against Gram-negative bacilli. Examples are cefotaxime, ceftriaxone, ceftazidime, aztreonam and cefpodoxime. These enzymes are not active against the cephamycins, and are inhibited by clavulanic acid.

4.0 WHAT TO TEST

4.1 Klebsiella pneumoniae and Escherichia coli most frequently are associated with ESBL production. The percentage of isolates expressing this phenotype is very variable although a recent study from the United States reported 83 ESBL-producing isolates from 906 consecutive isolates of Enterobacteriaceae over a 20 week period.⁽⁷⁾ ESBL-producing isolates of Enterobacter aerogenes, E. cloacae, Serratia marcescens, Morganella morganii, K. oxytoca, Citrobacter freundii, and C. koserii have been detected but appear to be relatively rare. The detection methods used for K. pneumoniae and E. coli have not been shown to be valid for other ESBL-producing bacteria.

4.2 When to test for ESBL production

4.2.1 All clinically significant isolates of E. coli or K. pneumoniae should be tested against beta-lactam drugs. Either the disk diffusion method, using the revised NCCLS interpretive criteria or an MIC method, where the drug concentration range extends to < 2 mg/L, may be used.

Although high level ESBL-producers are detectable using standard disk interpretive criteria⁹ or by the currently available automated methods, many others are detectable only by using the revised criteria.

Isolates that demonstrate reduced susceptibility to one or more of ceftazidime, cefuroxime, cefotaxime, ceftriaxone, cefpodoxime or aztreonam but remain susceptible to cefoxitin or cefotetan are considered as potential producers of ESBLs.

4.2.2 Isolates of E. coli or K. pneumoniae showing reduced susceptibility or resistance to the extended spectrum cephalosporins and to cefoxitin or cefotetan should be considered as potentially expressing AmpC resistance. Other resistance mechanisms, such as altered permeability, are not excluded at this stage.

4.2.3 The detection methods have not been validated for other Enterobacteriaceae.

4.2.4 Susceptibility testing using the revised criteria still may fail to detect low or relatively low expression of ESBL production. Additional testing to detect such production on a routine basis is not considered clinically necessary or cost-effective.

4.2.5 Selective testing for ESBL production should be considered for Gram negative enteric bacilli isolated from normally sterile body sites and where the infection may have been nosocomially acquired.

4.2.6 Selective additional testing should be considered for Gram negative enteric bacilli associated with a therapeutic failure when the original isolate of the same bacterium appeared susceptible to the extended spectrum cephalosporins.

5.0 METHODS OF TESTING

5.1 ESBL Detection Methods

International consensus guidelines on the detection of ESBLs have yet to be developed. In particular, there is no agreement on which isolates should be tested for these enzymes, what indicators should direct further testing, what methods should be used, and how the findings should be reported. This guideline provides Canadian laboratories with an interim national consensus document that will be modified or replaced by an agreed international standard of practice.

The detection methods fall into two main groups: those used to screen for ESBLs and those used to confirm that the observed reduction in susceptibility is due to an ESBL. Screening methods include both disk diffusion testing and minimum inhibitory concentration (MIC) methods. Phenotypic confirmatory testing currently rests on demonstrating an increase in the size of the inhibition zone around a disk containing an expanded spectrum cephalosporin, or a decrease in the MIC, when a beta-lactamase inhibitor (clavulanic acid) is added to the disk or MIC method. A >= 5 mm increase in a zone diameter or any reduction in the cephalosporin MIC of more than 3 doubling dilution steps, for the agent tested alone versus its MIC when tested in combination with clavulanic acid, indicates the possible presence of an ESBL.

5.2 Disk Diffusion Testing

In-vitro testing using the established NCCLS procedure and zone diameter breakpoints for Enterobacteriaceae may fail to detect intermediate susceptibility or resistance due to ESBL production. The revised guidelines,⁽⁸⁾ as shown in Table 1, have improved the detection sensitivity although it remains in the 80% range depending on the species. Only the cefpodoxime disk test approaches 100% sensitivity but it is not widely used in practice.

The table illustrates the "grey zone" for both zone diameters and MIC values that must be considered when a test isolate may be an ESBL-producing strain.

The use of a ceftazidime disk with a 5µg content was reported recently.⁽⁹⁾ The suggested zone diameter to detect ESBL production in K. pneumoniae was <= 17 mm and in E. coli <= 21 mm. The sensitivity of this method is of the same order as the revised NCCLS procedure using regular strength disks.

Guidelines for phenotypic confirmatory testing using ceftazidime (30µg) versus ceftazidime/clavulanic acid (30/10µg) and cefotaxime (30µg) versus cefotaxime/clavulanic acid (30/10µg) were issued by NCCLS in 1999.¹⁰ Regardless of the zone diameters given in Table 1, a >= 5 mm increase in a zone diameter for an antimicrobial agent tested in combination with clavulanic acid versus its zone when tested alone indicates probable ESBL production.

5.3 MIC Methods

Traditionally the automated rapid susceptibility testing methods have experienced difficulty in detecting beta-lactamases including ESBLs. The Vitek ESBL test has been reported as superior to the disk approximation test for detection of the enzymes.¹¹ A Microscan ESBL panel is under investigation. The potential of these automated systems will require the drug concentration range to extend below 2 mg/L.

A standard microdilution broth method to determine the MIC of ceftazidime and cefotaxime with and without 2 mg/L of clavulanic acid has been reported.¹⁰ NCCLS recommends 4 mg/L of clavulanic acid. The expected reduction in MIC in the presence of the beta-lactamase inhibitor is >= 3 log₂ dilution (3 doubling dilution) steps. This method seems to work well for ESBL-producing isolates of K. pneumoniae and E. coli but not with Enterobacter or Morganella isolates.

Detection and validation of ESBL-producing enteric bacilli by the Etest^® ESBL strip appears to be sensitive and reproducible.¹³ The strip carries two gradients: ceftazidime and ceftazidime with clavulanic acid. The test range with ceftazidime alone is 0.50 to 32 mg/L while in the presence of clavulanic acid it is 0.125 to 8 mg/L. The compression of the gradient does not compromise the accuracy of the MIC determination. Generally group 2be-expressing strains will show a reduction in the ceftazidime MIC of >= 5.5 to 7 log₂ (doubling) dilution steps but any reduction of >= 3 log₂ (doubling) dilution steps is considered as positive when using the Etest^® in a screening mode. It should be noted that not all ESBL producing strains are specific for ceftazidime. Strains with other substrate specificities may not be detected with the ceftazidime/clavulanic acid strip alone.

5.4 Disk Approximation or Double Disk Method

The disk approximation method uses either multiple target disks, or a single cefpodoxime disk, and a clavulanic acid disk. Disk content and the disk placement must follow a validated method.⁷

A Mueller-Hinton agar plate is inoculated with a suspension made from an overnight blood agar culture of the test strain as recommended for a standard disk diffusion susceptibility test. Disks containing the standard 30 µg of ceftazidime, or ceftriaxone, or aztreonam, or 10 µg of cefpodoxime are placed 15 mm apart (edge to edge) and from an amoxicillin-clavulanic acid disk containing 10 µg of the latter compound. The disk edge-to-edge distance recommended here is that reported⁷ as having greater sensitivity than the previous distance of 20 to 30 mm.

Following incubation for 16-20 hours at 35^o C, any enhancement of the zone of inhibition between a beta-lactam disk and that containing the beta-lactamase inhibitor is indicative of the presence of an ESBL.

Precise placement of the disks, correct storage of the clavulanate-containing disks, and performance of appropriate control tests are critical to the sensitivity of the disk approximation method.¹²

Caution is advised when testing organisms other than E. coli and K. pneumoniae due to lack of validation of the method.

5.5 Revised NCCLS Disk and MIC Methods 10

The quality control strains recommended for both disk diffusion and MIC testing are E. coli ATCC 25922 and K. pneumoniae ATCC 700603 (previously K6). On the initial screen, E. coli ATCC 25922 shows no growth and K. pneumoniae ATCC 700603 exhibits growth in broth dilution tests. On disk diffusion testing the zone diameter ranges for K. pneumoniae ATCC 700603 are cefpodoxime 9-16 mm, ceftazidime 10-18 mm, aztreonam 9-17 mm, cefotaxime 17-25 mm, ceftriaxone 16-24 mm. K. pneumoniae ATCC 700603 shows >= 3 dilution decrease in an MIC for an antimicrobial agent tested in combination with clavulanic acid versus its MIC when tested alone. In disk diffusion phenotypic confirmatory testing, E. coli ATCC 25922 exhibits <= 3 mm increase in zone diameter for an antimicrobial agent tested alone versus its zone when tested in combination with clavulanic acid. Under similar test conditions K. pneumoniae ATCC 700603 shows >= 5 mm increase in zone diameter.

5.6 Molecular Testing

Molecular methods to confirm or characterize ESBL production are available. Specific arrangements for the use of such testing should be made with the facility offering the service.

6.0 INTERPRETATION AND REPORTING

6.1 Clinical Significance of ESBL Production

The clinical outcome of patients infected with an ESBL-producing strain of enteric Gram-negative bacilli is uncertain. The medical literature fails to demonstrate that in vitro resistance to one or more extended spectrum cephalosporin drugs predicts clinical failure to the group. Patients with a good clinical outcome often were treated with an antibiotic combination or had a single organ/system infection such as a urinary tract infection or respiratory infection without evidence of pneumonia. Patients with a poor outcome usually had multi-organ/system infection or an associated life-threatening illness. It would seem prudent to err on the side of caution and consider that an ESBL-producing isolate may predict therapeutic failure with extended spectrum cephalosporin drugs.

6.2 Reporting

Patient reports must state that the isolate is a suspected or proven ESBL producer. A note should be included stating that ESBL production may predict therapeutic failure in some patients treated with drugs such as penicillins, ceftazidime, cefotaxime, ceftriaxone and aztreonam. Alternative therapy should be considered recognizing that ESBL producing strains may be resistant to other agents including the aminoglycosides.

6.3 Organisms sent to a reference laboratory should be stocked by the referring laboratory until conformation is received from the reference laboratory. The reference laboratory should stock the isolates for at least six months, or as required by provincial or local regulation.

6.4 For external/national surveillance reporting of isolates confirmed to be ESBL positive, report only one isolate per patient.

7.0 SUMMARY

Although observed for many years, there is increasing observation and detection of E. coli and Klebsiella as well as other Enterobacteriaceae that express ESBLs. The laboratory identification of ESBLs is still in an evolving state. A significant amount of work by many investigators has established the general guidelines iterated in this document for suspected and probable ESBLs for Klebsiella and E. coli. Laboratories can effectively use these guidelines in their own local situations to help monitor the emergence of ESBLs. The finding of such strains should prompt collaboration with reference laboratories to further understand the scope of this problem in Canadian laboratories.

8.0 REFERENCES

1. Sanders CC, Sanders WE Jr. ß-lactam resistance in gram-negative bacteria: global trends and clinical impact. Clin Infect Dis 1992;15:824-83

2. Livermore DM. ß-lactamases in Laboratory and Clinical Resistance. Clin Microbiol Rev 1995;8:557-584

3. Knothe GP, Shah P, Kremery V, Antai M, Mitsuhashi S. Transferable resistance to cefotaxime, cefoxitin, cefamndole and cefuroxime in clinical isolates of Klebsiella pneumoniae and Serratia marcescens. Infection 1983;11:315-317

4. Emery CL, Weymouth LA. Detection and Clinical Significance of Extended-Spectrum ß-Lactamases in a Tertiary-Care Medical Center. J Clin Microbiol 1997;35:2061-2067

5. Dubois SK, Marriott MS, Amyes SGB. TEM and SHV-derived extended spectrum ß-lactamases: relationship between selection, structure, and function. J Antimicrob Chemother. 1995;35:7-32

6. Katsanis GP, Spargo J, Ferraro MJ, Sutton L, Jacoby GA. Detection of Klebsiella pneumoniae and Escherichia coli strains producing extended-spectrum beta-lactamases. J Clin Microbiol 1994;32:691-696

7. Coudron PE, Moland ES, Sanders CC. Occurrence and Detection of Extended-Spectrum ß-lactamases in Members of the Family Enterobacteriaceae at a Veterans Medical Center: Seek and You May Find. J Clin Microbiol 1997;35:2593-2597

8. NCCLS. Performance Standards for Antimicrobial Susceptbility Testing; Eighth Informational Supplement. NCCLS document M100-S8. NCCLS, Wayne, PA 1998

9. Jacoby GA, Han P. Detection of extended-spectrum ß-lactamases in clinical isolates of Klebsiella pneumoniae and Escherichia coli. J Clin Microbiol 1996;34:908-911

10. NCCLS. Performance Standards for Antimicrobial Susceptbility Testing; Eighth Informational Supplement. NCCLS document M100-S9. NCCLS, Wayne, PA 1999

11. Sanders CC, Barry AL, Washington JA, Shubert C, Moland ES, Traczewski MM, Knapp C, Mulder R. Detection of extended-spectrum-beta-lactamse-producing members of the family Enterobacteriaceae with Vitek ESBL test. J Clin Micobiol 1996;34:2997-3001

12. Moland ES, Thompson KS. Extended-spectrum beta-lactamases of Enterobacteriaceae. J Antimicro Chemother 1994;33:666-668

13. Cormican MG, Marshall SA, Jones RN. Detection of Extended-Spectrum ß-lactamse (ESBL)-producing strains by the Etest ESBL Screen. J Clin Microbiol 1996;34:1880-1884

[CEQA-AGAR]