Comparison of Two Phenotypic Double Disc Diffusion Test (DDDT) for Detection of Extended Spectrum β-Lactamase Production by Enterobacteriaceae and Pseudomonas Species at a Tertiary Care Hospital in Bangladesh

Background: The detection of extended-spectrum β-lactamase-producing (ESBL) bacteria is of importance for infection control and epidemiological surveillance. Objective: The purpose of the present study was to compare two phenotypic methods for the detection of ESBL-positive Enterobacteriaceae and Pseudomonas species. Methodology: This cross sectional study was carried out in the Department of Microbiology at Mymensingh Medical College, Bangladesh from January 2011 to June 2011. Patients of all ages and genders presented with UTI or wound infection were taken as study population. Gram negative bacilli (GNB) were analyzed by two methods used for routine susceptibility testing which were Disk diffusion methods and MIC reduction methods. Two methods designed for the detection of ESBL production Ceftazidime and Ceftazidime plus Clavulinic acid, (CAZ/CAZC) and Cefotaxime and Cefotaxime plus Clavulinic Acid (CTX/CTXC) were used and the PCR was considered as gold standard for evaluation of the other test methods. Result: A total number of 300 GNB were isolated and identified of which 214(71%) were ESBL positive. For the disk diffusion method, resistant to third generation Cephalosporins were the highest 87.0%, when tested by ceftazidime and by MIC reduction methods were 67%. For the phenotypic confirmatory methods, specificities were 64% by (CAZ/CAZC), 59% by (CTX/CTXC) and by both method 52%. Among the phenotypic confirmatory ESBL positive strains by Genotypic method ESBL positive were 50.46% TEM, 18.69 % SHV and 46.72% CTX-M gene. Conclusion: Two-step strategies using both DDDT phenotypic methods are useful diagnostic tools for the detection of ESBL from the Gram negative bacilli.


Introduction
Antibiotic resistance has been emerged worldwide. It has been reported that the overuse of third generation of cephalosporins has been associated with the emergence of β-Lactamases 1 . These enzymes have serine at their active site and attack the amide bond in the β lactam ring of amtibiotics 2 . ESBLs are enzymes that mediate resistance to extended spectrum cephalosporins as well as monobactams 3 .
There are no definite guidelines for the detection of extended-spectrum beta-lactamase (ESBL) among Gram negative bacilli especially Enterobacteriaceae as well as Pseudomonas species. Though these species may cause hospital associated infection (HAI) [4][5][6] these are developed frequently multidrug resistant (MDR) [5][6] . These bacteria may constitute a reservoir for plasmid-mediated ESBLs transmission for other Enterobacteriaceae species [7][8][9] . Phenotypic detection of ESBLs in members of the Enterobacteriaceae co-expressing an AmpC beta-lactamase is complex as because AmpC expression may mask the synergy required for ESBL detection between third-generation cephalosporin and clavulanic acid. This problem may be circumvented by detection of ESBL by molecular technique with specific primer 10 . Most guidelines 20 Comparison of Two Phenotypic Double Disc Diffusion Test (DDDT) for Detection of Extended Spectrum β-Lactamase Production by Enterobacteriaceae and Pseudomonas Species at a Tertiary Care Hospital in Bangladesh recommend screening isolates based on decreased susceptibility to extended-spectrum cephalosporin in primary susceptibility testing followed by confirmation of ESBL production. However, still there is a controversy about the confirmation of ESBL 11 . Therefore, detection of ESBL-producing organisms has become a concern for general hospitals and private practice laboratories. The current changes in clinical MIC breakpoints for extended-spectrum cephalosporins and for aztreonam against Enterobacteriaceae by EUCAST and CLSI decrease the likelihood of interpreting an ESBL-producing Enterobacteriaceae as susceptible to extended-spectrum cephalosporins [12][13][14] . Several phenotypic methods have been developed to confirm ESBL production by Enterobacteriaceae 15-17 . Therefore the purpose of the present study was to compare two phenotypic methods for the detection of ESBL-positive Enterobacteriaceae and Pseudomonas species.

Methodology
This cross sectional study was carried out in the Department of Microbiology at Mymensingh Medical College, Mymensingh, Bangladesh from January 2011 to June 2011. Mymensingh is 200 KM away from the capital of Bangladesh which is the largest district of this country and Mymensingh Medical College is the only teaching hospital with more than 500 beds with emergency as well as OPD facilities. Patients presented with UTI or wound infection were at any age with both sexes were taken as study population. Non-repetitive clinical isolates were collected from MMCH both the outpatients and inpatients department over a period of 6 months. Urine and wound infection were used as specimen. Laboratory work was carried out in the department of Microbiology in Mymensingh Medical College, Bangladesh. Specimens were collected aseptically 18 . All samples were routinely cultured on MacConkey and blood agar plates at 37 0 C aerobically for 18 hours. Gram negative isolates were further characterized by standard biochemical tests 19 .The susceptibility to antibiotics was determined by Kirby Bauer method according to CLSI 2010 protocols for Gram negative panels 20 . E. coli ATCC 25922 was used as control strains. Three diagnostic tests for ESBL production were performed which were screening test for ESBL by disc diffusion test (DDT) and MIC reduction method, confirmatory test by double disc diffusion test (DDDT) and multiplex PCR for TEM, SHV and CTX-M genes. Isolates were screened for ESBL production by using disc diffusion test on Muller Hinton agar according to the CLSI 21 recommendationswhere isolates showing inhibition zone size of ≥22mm with ceftazidime (30μg), ≥25mm with ceftriaxone (30μg), ≥ 27mm with cefotaxime (30μg), ≥27 mm with Aztreonam (30 µg) were suspected for ESBL production. MIC reduction method ≥8µgm/ml with Ceftazidime was performed. E. coli ATCC 25922 was used as a negative control. Phenotypic confirmatory test for ESBL producers were done by double disc diffusion test (DDDT) by two methods which were eftazidime and Ceftazidime plus Clavulinic acid (CAZ/CAZC) and Cefotaxime and Cefotaxime plus Clavulinic Acid (CTX/CTXC). Plasmid DNA was isolated from bacterial cells by alkaline lysis method by Medici et al 22 . Supernatant 1µl was used as template. The PCR primers and cycling conditions used were previously described [23][24][25] . Primers used for TEM gene were TEM F, TEM R, SHV gene were SHV F, SHV R, CTX-M gene were CTX-MU-1, CTX-MU-2. Amplification was performed in an Thermocycler (Eppendrof, Takara, Japan) in 1.0% agarose gel and visualized by staining with ethidium bromide (0.5 mg/ml) running through horizontal electrophoresis. A 100 bp ladder molecular weight marker (Roche, USA) was used to measure the molecular weights of amplified products. The images of ethidium bromide stained DNA bands were digitized using a gel documentation system (AlphaimagerTM 3400, USA) by UV light. The qualitative data were expressed as frequency and percentage. The comparisons of qualitative data were performed by Chi-square test if data were normally distributed. Statistical analysis was performed by using window based computer software devised with Statistical Packages for Social Sciences (SPSS-17) (SPSS Inc, Chicago, IL, USA). 95% confidence limit was taken. The summarized data was interpreted accordingly and was then presented in the form of tables. P value less than 0.05 was taken as level of significance.

Discussion
Detection of ESBLs are a great challenge in the clinical microbiology laboratory 21-25 . The CLSI guidelines regarding the sensitivity breakpoints designated for Klebsiella species and Escherichia coli against cefotaxime, ceftriaxone, and ceftazidime are inefficient for the detection of ESBLs [26][27] . Failure to detect ESBL production by routine disc-diffusion tests has been well documented 26 . Highly informative studies were performed by Tenover et al 26 in the United States and Livermore et al 28 in Europe and were reported that ESBL mediated resistance are frequently encountered difficulties with both automated and disk diffusion methods. This might be due to the variable affinity of enzymes for different substrates and inoculum effect 29 . Furthermore, ESBL-producing organism gives the treatment failure with third generation cephalosporin if the infection is outside the urinary tract 27 . In addition to that reports of failures in both animal models and clinical settings are well documented when third-generation cephalosporins are used to treat ESBL infections 30-31 . The sensitivity of any particular extended-spectrum cephalosporin used as a screening test can be dependent on the geographic variation and the resistance patterns of these organisms 20 . More resistant to ceftazidime is reported and SHV-3 like ESBLs has been found which is 100.0% resistant to ceftriaxone; whereas only 50.0% are resistant to ceftazidime 30 . In this study resistant to Ceftazidime is found in 87.0%. Moland et al 31 has used ceftazidime resistance as marker for ESBL production which is 78.0% in a broth dilution format. Similarly it has been found statistically more significant by these combinations. Though ceftazidime plus clavulinic acid (CAZ/CAZC) is the best single disc diffusion test for the detection of ESBL 25 , it cannot cover all the strain. Use of only one combination may fail to detect ESBL positive strains; thus this might give underestimation of prevalence of ESBL. In this study both combinations were found statistically significant. The use of cefotaxime, ceftazidime and ceftriaxone as the only indicator of ESBLs screening can no longer be recommended. If only one indicator antibiotic would be used for screening,

22
Comparison of Two Phenotypic Double Disc Diffusion Test (DDDT) for Detection of Extended Spectrum β-Lactamase Production by Enterobacteriaceae and Pseudomonas Species at a Tertiary Care Hospital in Bangladesh cefpodoxime has proven to be the best molecule for screening all types ESBLs producers in clinical sample and this also produces AmpC β-lactamase 23 . In this study two combinations with clavulanic acid like CAZ/CAZC and CTX/CTXC are tested and have been found that E. coli and Klebsiella species showed maximum ESBLs production in CAZ/CAZC combination which correlates with other studies 31 . However, the combination of ceftazidime and cefotaxime still missed two strains producing the SHV-5 and SHV-7 ESBLs 30 . If the both combination does not use the actual figure may not appear and may show low prevalence rate. These types of discrepancies between susceptibility data and disc diffusion results have increased the need for an improved method of ESBL detection which can be incorporated into the routine susceptibility procedure. This study should be viewed in the context of several limitations. Firstly, because of constraints of time and budget, this was a single centered with relatively small sample size; therefore, the results cannot be generalized to the whole country. Furthermore, antibiotic history of the patients before and after in vitro sensitivity test was not considered and only in vitro drug response was taken as the result. PCR-based molecular assays have their own limitations. These need expensive equipment and reagents and expert personnel which are not always available to the diagnostic laboratory. In addition, primers have to be designed within low mutation rate regions and the use of specific primers hinders the identification of novel resistance genes giving false negative results. However, so far this type of study has not been performed in Bangladesh. Further study is needed to correlate between in vitro and in vivo drug response among the patients infected with ESBLs producing bacteria.

Conclusions
In conclusion routine used of both DDDT combinations is useful for the detection of ESBL in clinical samples. This is a better method for the detection of ESBL producing bacteria especially enterobacteriaceae and Pseudomonas species.