Background & objectives: Identifying organisms that harbour extended spectrum beta lactamases (ESBLs) is a major challenge for a diagnostic clinical microbiology laboratory. Wide variety of ESBLs produced and lack of a sensitive phenotypic method for their detection make the detection of ESBLs difficult and is responsible for under-recognition. The present study was undertaken to evaluate phenotypic characteristics, initial screening tests and established confirmatory phenotypic methods for detection of ESBLs Klebsiella pneumoniae isolates prevalent in a hospital in north India.
Methods: One hundred, non-repeat clinical isolates of K. pneumoniae collected over a period of six months were included in the study. Susceptibilities of the isolates to 20 different antimicrobial agents were determined. Agar dilution and broth dilution methods were used to determine minimum inhibitory concentrations (MICs) of ceftazidime (CAZ) and cefotaxime (CTX). CAZ and CTX were used with and without clavulanic acid to detect ESBL harbouring isolates. Using agar dilution and broth dilution, MIC reduction of two and three doubling dilutions were evaluated as a criterion for ESBL harbouring isolates. Standard double disk synergy test (DDST) with disks placed at 30mm and modified DDST with disks placed at 16mm center-to-center distance, using at least two different third generation cephalosporins and combined disk method were also performed to detect ESBL harbouring isolates.
Results: Multi-drug resistance (resistance to three or more antimicrobials of different classes) was found among 94 per cent of the isolates. Pooling the results of all the three confirmatory techniques MIC reduction of > 3 doubling dilutions using broth dilution method (using CTX and CAZ), combined disk method [(using CTX, ceftriaxone [(CRO), CAZ and aztreonam)] and standard DDST (using CTX, CRO, CAZ and aztreonam), revealed as many as 87 per cent of the isolates as ESBL producers. CTX had greater sensitivity in identifying isolates that harboured ESBLs. Modified DDST using CTX was as sensitive method as broth dilution method and combined disk method in detecting ESBL harbouring isolates. MIC reduction technique using agar dilution method and standard DDST had lowest overall sensitivity in detecting ESBLs.
Interpretation & conclusion: Modified DDST using at least two different third generation cephalosporins was considered to be the best technique for detection of ESBL producing K. pneumoniae at our hospital. MIC reduction test with > 2 doubling dilution reduction in MICs was found to be a better criterion than the presently recommended > 3 doubling dilution reduction. For screening of potential ESBL producers, MIC determination using agar dilution was as good as that using broth dilution method. However, while performing MIC reduction test agar dilution method was found highly unreliable for detection of ESBL harbouring isolates.
Key words Double disk synergy test - ESBLs - K. pneumoniae - MIC reduction test
Despite the fact that there are large numbers of Gram negative clinical isolates that harbour extended spectrum beta lactamases (ESBLs), very few clinical laboratories routinely perform ESBL detection, especially in developing countries, due to costs involved, work load and extra labour. Efforts have been made for detection and interpretation of ESBLs among clinical isolates and various cut off points of disk diffusion and minimum inhibitory concentrations (MICs) have been described1-3.
A key problem in the detection of ESBL producers is the possibility of low level expression of enzyme and of inoculum effect, resulting in variable MIC values and zone diameters by disk diffusion testing. Although different enzymes of TEM- and SHV- class show some patterns in antibiogram, the enzymes differ in the level of resistance conferred to different antimicrobial compounds. Therefore, specific detection methods such as the double disk synergy test (DDST), combined disk method, E-test and three-dimensional methods have been described1,4-6. Different studies have reported greater sensitivity or reliability of detection of one method over the other as well as the success of particular antimicrobial substrates. Of these methods, it is still impractical to use E test for detection of ESBLs on routine basis in diagnostic laboratories because of forbidden cost in developing countries. Moreover, difficulty in interpretation of E test has been reported in some cases due to 'phantom effect' 7,8. Three-dimensional test is still evolving9. In an earlier study different phenotypic techniques using well-characterised ESBL harbouring isolates have been compared and the investigators have emphasized to extend similar studies to the various clinical isolates8.
Detection rates for a particular method generally vary with the types of ESBLs produced and the geographical area where these clinical isolates are prevalent. Therefore, standardization of the techniques needs to be done to exploit the method for maximal detection of ESBL producers without false positive or negative results. Many studies from India have reported occurrence of ESBLs in various members of Enterobacteriaceae using different phenotypic methods9-19. The present study was carried out to evaluate phenotypic characteristics of clinical isolates of Klebsiella pneumoniae in a tertiary care hospital in north India and to determine the best phenotypic technique for detection of ESBLs among the clinical isolates.
Material & Methods
One hundred consecutive, non-repeat clinical isolates of K. pneumoniae obtained from various clinical specimens sent to microbiology laboratory at University College of Medical Sciences and Guru Tegh Bahadur Hospital, Delhi during March and August 2001 were included in the study. The isolates were identified with standard biotyping methods20. Of these neonatal intensive care unit (NICU) contributed to 19 isolates, and 31 isolates each were recovered from wound swabs from burns unit, and samples from various sites in patients admitted to inpatient units. Samples from outpatients contributed 18 isolates. One isolate was recovered from a patient in ICU.
Antimicrobial susceptibilities of the isolates to 20 different antimicrobial agents (µg) viz. cefoperazone, (75) ceftazidime (CAZ), (30) cefotaxime (CTX) (30) aztreonam (ATM) (30) ceftriaxone (CRO) (30) cefepime, (30), cefoxitin (30), cefotetan, (30) cotrimoxazole (1.25/23.75) amikacin, (30) gentamicin, (10) kanamycin, (30) ciprofloxacin, (5) chloramphenicol, (30) pipracillin, (100) pipracillin + tazobactam, (100/10) amoxycillin + clavulanate, (20/10), loracarbef (30) imipenem (10) and netilmicin (30) were determined by standard disk diffusion (SDD) using commercially available disks (Oxoid, Basingstoke, UK) and were categorized as sensitive, intermediate and resistant as per National Committee for Clinical Laboratory Standards (NCCLS) guidelines wherever applicable1. Quality control was achieved by using standard strain of Escherichia coli ATCC 25922. NCCLS recommendations to identify potential ESBL producing isolates using standard disk diffusion technique were followed, i.e., zone sizes of
MICs of CAZ and CTX for these isolates were determined by the agar dilution and broth dilution methods using series of dilutions containing the antimicrobial (CAZ or CTX) alone (final concentrations of 512 - 0.0625µg/ml), and with 4µg/ml clavulanic acid (CA; Ranbaxy, India). MIC cut-off points for ESBL producers were taken according to the NCCLS recommendations, isolates with MICs > 2µ/ml for CAZ, and MICs > 8µg/ml for CTX were identified as potential ESBL producers1. The MIC reduction of two times and three times doubling dilutions were assessed for cut-off to be taken as ESBL phenotype.
Standard and modified double disk synergy test (DDST) was performed using disks of 30 µg each of CAZ and CTX along with AMCA (amoxycillin 20 µg x clavulanic acid 10 µg). The disks were placed at the distance 30 and 16 mm from each other (centre to centre) and incubated at 37°C overnight12. The organism was considered habouring ESBLs if the zone of inhibition around the CAZ and/or CTX disk showed a clear cut increase towards the AMCA disk4,12.
Combined disk method was used as recommended by the NCCLS1. One disk each of CAZ/CA (30 /10µg) and CTX/CA (30/10µg) were applied to the surface of the inoculated plate. A disk each of CAZ and CTX were also applied. A 5mm increase in zone diameter for either antimicrobial agent tested in combination with CA versus its zone when tested alone was taken as an indication of ESBL producing isolates. Quality control was achieved using E. coli ATCC 25922 and K. pneumoniae ATCC 700603.
Criteria for confirmed ESBL phenotype: When the isolate was found to harbour ESBL by atleast two of the three confirmatory criteria that included MIC reduction of > 3 doubling dilutions using broth dilution method (using CTX and CAZ), combined disk method (using CTX, CRO and CAZ), and DDST (using CTX, CRO and CAZ) with 30 mm distance center-to-center between the two disks, it was labeled as confirmed ESBL phenotype.
Results & Discussion
All the isolates were found susceptible to imipenem (Table I). Multi drug resistance (resistance to three or more antimicrobials of different classes) was found among 94 per cent of the isolates. Six isolates that were not multi drug resistant were all ESBL non-producers. Also, 59 of the total isolates were resistant to all the tested third generation cephalosporins (3GC) and aztreonam. Among non-beta-lactam agents, chloramphenicol and co-trimoxazole were the two antimicrobials against which maximum ESBL harbouring isolates were found susceptible (Table II). Susceptibility to aminoglycosides was found low among the ESBL producers ranging between 6-20 per cent. Among all the isolates studied 37.9 per cent were found resistant to the piperacillin+tazobactam combination while all isolates were found susceptible to carbapenems studied. The reasons for this pattern are not known. None of the ESBL harbouring isolates was sensitive to combination of amoxycillin and clavulanic acid (Table II). An association was found between aztreonam and gentamcin resistance and ESBL production. Isolates that were resistant to gentamicin and aztreonam harboured ESBLs but not all the ESBL harbouring isolates were resistant to the two antimicrobials. No other association between antimicrobial susceptibility and ESBL production could be identified.
Using standard disk diffusion (SDD) as screening method for identifying potential ESBL producers, ceftriaxone was the most efficient antimicrobial in screening isolates as potential ESBL producers followed by CTX (Table III). Ceftriaxone missed three ESBL harbouring isolates. Cefotaxime and aztreonam missed seven and nine ESBL harbouring isolates respectively. Ceftazidime was the least efficient among these four antimicrobials missing 14 isolates that were later identified as harbouring ESBLs.
MICs determined using agar dilution and broth dilution methods revealed similar results. The detailed distribution of MICs of the isolates for ceftazidime and cefotaxime is shown in Fig.1 and 2. Among ESBL harbouring isolates MICs of
ESBL producing K. pneumoniae may have higher MICs against antimicrobials compared with those of ESBL non-producing isolates. Since these elevated MICs to CAZ do not reach the NCCLS recommended break point levels for potential ESBL producers (>2µg/ml), therefore, such isolates may be incorrectly dismissed as ESBL non-producing isolates1. The isolates having high MICs of oxyimino beta-lactam agents can be labeled as potential ESBL producers. In the present study, nine isolates could not meet the NCCLS criteria for potential ESBL producers. Moreover, MICs of four ESBL harbouring isolates were
Using MIC reduction test as confirmatory test for detection of ESBL producing isolates, with > 3 doubling dilution MIC reduction criterion for ESBL producers, CTX detected greater number of ESBL harbouring isolates than CAZ. However, use of > 2 doubling dilution MIC reduction criterion increased the sensitivity of detection of ESBL harbouring isolates (Table IV). The extra isolates that were detected by modified criterion (> 2 doubling dilution MICs reduction) were also confirmed as harbouring ESBLs using combined disk method and standard double disk synergy test.
When standard DDST (30mm distance center-to-center) was used as confirmatory test for detection of ESBL producing isolates, maximum of 22 isolates were detected as ESBL producers using CRO. However, reducing the distance to 16mm center-to-center markedly increased the sensitivity of the test (Table IV). CTX and CRO detected maximum number of ESBL harbouring isolates followed by aztreonam and ceftazidime. The results obtained by the combined disk method revealed similar results CTX and CRO were better antimicrobials for detection of ESBL harbouring isolates than aztreonam and CAZ.
Pooling the results of all the three confirmatory techniques MlC reduction of > 3 doubling dilutions using broth dilution method (using CTX and CAZ), combined disk method (using CTX, CRO, CAZ and aztreonam), and DDST (using CTX, CRO, CAZ and aztreonam) with 30 mm distance center-to-center between the two disks, revealed as many as per 87 per cent of the isolates as ESBL producers. CTX had greater sensitivity in identifying isolates that harboured ESBLs. Modified DDST using CTX was as sensitive as broth dilution method and combined disk method in detecting ESBL harbouring isolates. MIC reduction technique using agar dilution method and standard DDST (30mm distance center-to-center) had lowest overall sensitivity in detecting ESBLs.
The distance between disks is critical for each enzyme, as the test depends on the concentration of both beta-lactam antibiotic and inhibitor. Several attempts may be required to detect an ESBL producer. Several modifications, including the choice of drugs tested and varying the distance between the disks, have been recommended4-6. We tried different distances viz., 30, 20, 16 and 14mm for detection of ESBL harbouring isolates. Best results were obtained by using at 16mm distance center-to-center between the two disks (unpublished data). Increasing the distance further between the two disks resulted in missing a large number of ESBL harbouring isolates. On the other hand, decreasing the distance further interfered with the interpretation of the results. A study from India also used 16 mm distance and reported good results12. Several workers from India and other countries have used 30, 25, 21 and 20mm distance between the disks and few have labeled the method as inadequate for detecting all the ESBL producing isolates4,5,11,24,25. In the present study, use of cefotaxime and ceftriaxone in the DDST resulted in the detection of a larger number of ESBLs harbouring isolates than ceftazidime. The results were similar to the combined disk method currently recommended by the NCCLS. Thus, for laboratories that perform susceptibility testing by disk diffusion, modified DDST could be easily incorporated into an already existing system. It has the added benefit that there is no need to measure zone sizes hence removing the subjectivity and is easy to read by recording presence or absence of synergy. Further, it requires no extra time in setting up the test and reading it.
MIC reduction of > 2 or > 4 fold has been used in the past6,14,26. It is important to note that three additional isolates identified by using > 2 doubling dilution reduction criterion were later confirmed by standard DDST and combined disk method as ESBL producers. Further studies are needed to confirm these findings.
Our findings showed that modified DDST has an added advantage over the combined method since it is easier to perform and quick to interpret. In our study, the modified DDST was found as good as MIC reduction test using broth dilution in detection of ESBL producers. A study from Turkey has reported MIC reduction as a better technique than standard DDST for the detection of ESBL harbouring isolates27.
It was observed that when used alone, CAZ was able to detect lesser number of the ESBL producing isolates as compared to CTX by all the three confirmatory techniques. Combination of two or more techniques increased the detection rate of ESBL producers. It is important to note that for screening of potential ESBL producers MIC determination using broth dilution was as good as that using agar dilution method. However, while performing MIC reduction test agar dilution method was found highly unreliable for detection of ESBL harbouring isolates. The material, labour and cost involved may preclude the use of MIC reduction test for detection of ESBL harbouring isolates in routine clinical laboratories.
To conclude, modified DDST using at least two different 3GCs was the best technique for detection of ESBL producing K. pneumoniae at our hospital. MIC reduction test with ≥ 2 doubling dilution reduction in MICs was found to be a better criterion than the presently recommended ≥ 3 doubling dilution reduction. Our observations suggested that additional testing to detect ESBL production in the clinical isolates on a routine basis, would be necessary to institute appropriate antibiotic therapy.
1. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial susceptibility testing. 10th Informational Supplement, Wayne, Pennsylvania: National Committee for Clinical Laboratory Standards Document M100-S10; 2000.
2. D' Agata E, Venkataraman L, DeGirolami P, Weigel L, Samore M, Tenovar F. The molecular and clinical epidemiology of enterobacteriaceae-producing extended-spectrum beta-lactamase in a tertiary care hospital J Infect 1998; 36: 279-85.
3. Dandekar PK, Barrett NL, Nightingale CH, Nicolau DP. Utilization of extended-spectrum beta-lactamase (ESBL) detection systems in microbiology laboratories: survey of Connecticut hospitals from 1998-2002. Conn Med 2003; 67 : 149-52.
4. Jarlier V, Nicolas MH, Fournier G, Philippon A. Extended broad-spectrum beta-lactamases conferring transferable resistance to newer beta-lactam agents in Enterobacteriaceae: hospital prevalence and susceptibility patterns. Rev Infect Dis 1988; 10 : 867-78.
5. Thomson KS, Sanders CC. Detection of extended spectrum beta-lactamases in members of the family Enterobacteriaceae: comparison of the double-disk and three-dimensional tests. Antimicrob Agents Chemother 1992; 36 : 1877-82.
6. Cormican MG, Marshall SA, Jones RN. Detection of extended-spectrum beta-lactamase (ESBL)-producing strains by the E test ESBL screen. J Clin Microbiol 1996; 34 : 1880-4.
7. Bradford PA. Extended spectrum beta-lactamases in the 21st century: characterisation, epidemiology, and detection of this important resistance threat. CHn Microbiol Rev 2001; 14 : 933-51.
8. MacKenzie FM, Miller CA, Gould IM. Comparison of screening methods for TEM- and SHV- derived extended - spectrum beta-lactamase detection. Clin Microbiol Infect 2002; 8 : 715-24.
9. Manchanda V, Singh NP. Occurrence and detection of AmpC beta-lactamases among Gram-negative clinical isolates using a modified three-dimensional test at Guru Tegh Bahadur Hospital, Delhi, India. J Antimicrob Chemother 2003; 5 : 415-8.
10. Jain A, Roy I. Gupta MK, Kumar M, Agarwal SK. Prevalence of extended-spectrum beta-lactamase-producing Gram-negative bacteria in septicaemic neonates in a tertiary care hospital. J Med Microbiol 2003; 421-5.
11. Subha A, Ananthan S. Extended spectrum beta-lactamase mediated resistance to third generation cephalosporins among Klebsiella pneumoniae in Chennai. Indian J Med Microbiol 2002; 20 : 92-5.
12. Ananthakrishanan AN, Kanungo R, Kumar A, Badrinath S. Detection of extended spectrum beta-lactamase producers among surgical wound infectons and burns patiens in JIPMER. Indian J Med Microbiol 2000; 18 : 160-5.
13. Hansotia JB, Agarwal V, Pathak AA, Saoji A M. Extended spectrum beta-lactamase mediated resistance to third generation cephalosporins in Klebsiella pneumoniae in Nagpur, central India. Indian J Med Res 1997; 105 : 158-61.
14. Abigail S, Mathai E, Jesudason MV, John TJ. Ceftazidime resistance among Klebsiella pneumoniae in south India Indian J Med Res 1995; 102 : 53-5.
15. Shukla I, Tiwari R, Agrawal M. Prevalence of extended spectrum beta-lactamase producing Klebsiella pneumoniae in a tertiary care hospital Indian J Med Microbiol 2004; 22 : 87-91.
16. Revathi G. Singh S, Singh S. Detection of expanded spectrum cephalosporin resistance due to inducible lactamases in hospital isolates. Indian J Med Microbiol 1997: 15 : 113-5.
17. Mathur P. Kapil A. Das B. Dhawan B. Prevalence of extended spectrum beta lactamase producing Gram negative bacteria in a tertiary care hospital. Indian J Med Res 2002; 115 : 153-7.
18. Khurana S, Taneja N, Sharma M. Extended spectrum beta-lactamase mediated resistance in urinary tract isolates of family Enterobacteriaceae. Indian J Med Res 2002; 116 : 145-9.
19. Shahid M. Malik A, Agrawal M, Singhal S. Phenotypic detection of extended-spectrum and AmpC beta-lactamases by a new spot-inculation method and modified three-dimensional extract test: comparison with the conventional three-dimensional extract test J Antimicrob Chemother 2004; 54 : 684-7.
20. Crichton PB. Enterobacteriaceae: Escherichia, Klebsiella, Proteus and other genera. In: Collee, I G, Fraser A G, Marmion, B P, Simmons A, editors. Mackie & McCartney practical medical microbiology, 14th ed. Edinburgh, UK: Churchill Livingstone. 1996: p. 361-84.
21. Jorgensen JH, Turnidge JD. Susceptibility testing methods: Dilution and disk diffusion methods. In: Murray PR, Baron EJ, Pfaller MA, Jorgenson JH, Yolken RH, editors. Manual of clinical microbiology, 8th ed VA, USA: ASM Press 2003; p. 1108-27.
22. West PW. Extended spectrum beta-lactamase producing Klebsiella spp. Br J Biomed Sci 2000; 57 : 226-33.
23. Yuan M, Aucken H, Hall LM. Pitt TL, Livermore DM. Epidemiological typing of klebsiellae with extended-spectrum beta-lactamases from European intensive care units. J Antimicrob Chemother 1998; 41 : 527-39.
24. Emery CL. Weymouth LA. Detection and clinical significance of extended-spectrum β-lactamases in a tertiary-care medical center. J Clin Microbiol 1997; 35 : 2061-7.
25. Coudron PE, Moland ES, Sanders CC. Occurrence and detection of extended-spectrum beta-lactamases in members of the family Enterobacteriaceae at a Veterans medical center: seek and you may find. J Clin Microbiol 1997; 35 : 2593-7.
26. Ho PL, Chow KH, Yuen KY, Ng WS, Chau PY. Comparison of a novel, inhibitor-potentiated disk-diffusion test with other methods for the detection of extended-spectrum beta-lactamases in Escherichia coli and Klebsiella pneuinoniae. J Antimicrob Chemother 1998; 42 : 49-54.
27. Kocazeybek BS. Antimicrobial resistance surveillance of Gramnegative isolated from intensive care units of four different hospitals in Turkey. Evaluation of the prevalence of extended-spectrum and inducible beta-lactamases using different E-test strips and direct induction methods. Chemotherapy 2001; 47 : 396-408.
Vikas Manchanda, N.P. Singh, R. Goyal, A. Kumar & S.S. Thukral*
Department of Microbiology, University College of Medical Sciences & Guru Tegh Bahadur Hospital & *V.P. Chest Institute, University of Delhi, Delhi, India
Received August 12, 2004
Reprint requests: Dr N.P. Singh. Department of Microbiology, University College of Medical Science & Guru Tegh Bahadur Hospital
Dilshad Garden, Delhi 110095, India
Copyright Indian Council of Medical Research Oct 2005
Provided by ProQuest Information and Learning Company. All rights Reserved