Background: Antimicrobial resistance mechanisms including the beta-lactamases (βLs) such as extended-spectrum beta-lactamases (ESBLs), AmpC βLs, metallo-beta-lactamases (MBLs) and inhibitor-resistant TEM βLs (IRTs) are continuously developing, and infections due to such resistance bacteria are left with limited treatment options. Aims: This study was conducted to evaluate the activity of different beta-lactams and non-beta-lactam antibiotics against βL producing Gram-negative clinical isolates. Materials and Methods: Isolation, identification of different βLs and antibiotic susceptibility testing in 350 non-repeat, consecutive, Gram-negative clinical isolates were performed using the automated Vitek 2 system (bioMe΄rieux, Marcy l’Etoile, France). Eighteen antimicrobials were included in the AST N-280 panel of Vitek 2. Phenotypic disk diffusion methods were also employed for the detection of various βLs. Result: Both the automated and phenotypic methods identified 34% (119/350) βLs of which 18.29%, 6%, 1.14%, 2.86%, 3.42%, and 1.43% were ESBLs, AmpC βLs, MBLs, ESBL with AmpC, ESBL with MBL and AmpC with MBL, respectively. IRTS were detected in 3 (0.86%) isolates. Among the β-lactams and β-lactams with βL inhibitors, the least resistant antimicrobials against βLs were respectively cefepime (73%) and pipercillin/tazobactam (45.5%). Regarding the non-beta-lactams, maximum sensitivity was observed with colistin (96.4%), followed by tigecycline (94%), meropenem (87.4%), amikacin (86.6%) and ertapenem or imipenem (82.4%). The activity of nitrofurantoin was relatively good with a sensitivity of 61.4%. Conclusion: In our setting, colistin, tigecycline and meropenem turned out to be the best available antimicrobials to tackle the infections caused by βLs. Nitrofurantoin could be used in the management of urinary tract infection caused by multidrug-resistant isolates if its activity is high enough.
Keywords: Beta-lactamases, gram-negative isolates, Vitek 2 system
Beta-lactamases (βLs) are enzymes that hydrolyze the β-lactam ring of β lactam antibiotics such as penicillins and cephalosporins. The predominant mechanism for resistance to β-lactam antibiotics in Gram-negative bacteria is the synthesis of βLs.  Three distinguished major groups of these enzymes are class C cephalosporinases (AmpC), extended-spectrum beta-lactamases (ESBLs) and carbapenemases including class B metallo-beta-lactamases (MBLs).  Co-production of >1 type of βL poses a greater threat.  It is a challenge to treat such cases and could have enormous consequences if left undetected and untreated.
Studies suggest that the use of cephalosporins, including cephamycins and cefepime for ESBLs, is associated with a worse outcome, despite apparent in vitro susceptibility.  Despite the fact that the hydrolysis rates for cefepime and cefpirome by the AmpC enzymes are very low, and the estimated km values of these antibiotics are high, reflecting lower enzyme affinity, it is unsettled whether they can be used.  βL inhibitors specially tazobactam remain active against ESBL isolates, but they have much less effect on AmpC enzymes, although some are inhibited by tazobactam or sulbactam. , The only beta-lactams to retain activity against ESBL or AmpC-producing organisms are the carbapenems. 
The high rates of concurrent resistance to aminoglycosides, fluoroquinolones, and trimethoprim-sulfamethoxazole even after documentation of in vitro activity and the potential for emergence of resistance on treatment often preclude their use for empirical coverage. 
Although tigecycline and colistin have been used successfully against βL producing organisms, more recent data suggest that the resistance to colistin is emerging.  Nitrofurantoin has been used successfully for a long time for the prophylaxis and treatment of acute lower urinary tract infections in adults, children and pregnant women, but the increased emergence of antibiotic resistance has made nitrofurantoin a suitable candidate for the treatment of infections caused by multidrug-resistant pathogens. 
The present study was undertaken to survey the activity of β-lactam and non-β-lactam antibiotics against βLs producing Gram-negative clinical isolates using an automated Vitek 2 system, and to assess the levels of coresistance and alternative therapeutic options available in our set up.
The study included a total of 350 consecutive non-duplicate Gram-negative isolates obtained over a period of 1-year (July, 2012 to June, 2013) from different clinical specimens such as urine (n = 120), blood (n = 60), pus (n = 35), wound swab (n = 50), sputum (n = 31), intravenous catheters (n = 19), stool (n = 15), vaginal swab (n = 11), peritoneal fluid (n = 3) or pericardial fluid (n = 6) from the out-patient and in-patient departments of medicine, surgery, obstetrics and gynecology, orthopedic, and pediatrics units. It covered the patients of all age groups and both sexes.
Identification and antibiotic sensitivity pattern of the causative bacteria from the various clinical samples were performed by the Vitek 2 system (bioMe΄rieux, Marcy l’Etoile, France) following presence of pure culture of isolated colonies of the microorganisms on blood agar and MacConkey’s agar.  Following the manufacturer’s instructions, they were inoculated onto the identification and the AST-N280 cards, on which different concentrations of 18 antibiotics were assayed. This system also features an Advanced Expert System that interprets the antibiotic resistance patterns, validates the results and reports the resistance phenotype, and has proved useful in calculating MIC values.
Phenotypic methods such as combined disk test using disks containing 30 μg of cefotaxime, ceftazidime or cefepime and disks containing a combination of the three drugs plus 10 μg of clavulanic acid for ESBL production, AmpC disk test for AmpC βLs and combined disk or disk enhancement test using imipenem and imipenem with ethylenediaminetetraacetic acid for MBLs were employed following standard protocols. ,,
For the detection of the co-existence phenotype of both ESBL and AmpC, the original double disk synergy test was modified by placing a disk of piperacillin-tazobactam (100/10 μg) at a distance of 20 mm from cefepime (30 μg) disk. 
Quality control was achieved using Klebsiella pneumoniae (ATCC 700603), Escherichia More Details coli (ATCC 25922) and known AmpC positive E. coli. Reference strains of carbapenemases positive K. pneumoniae ATCC BAA-1705 and negative K. pneumoniae ATCC BAA-1706 were also included in the study.
Chi-square test was used to see the significance between the resistance level of various drugs in βL producers and non-producers using SPSS 20 software (SPSS South Asia Pvt Ltd. No.2353/1-4, 4 th Floor, Dolphin, Hennur, Bangalore, India). Where the cell frequency was <5, Fisher’s exact test was applied. P ≤ 0.05 was considered as significant.
The Vitek 2 system isolated E. coli (n = 91), K. pneumoniae (n = 73), Citrobacter freundii (n = 19), Enterobacter cloacae (n = 18), Proteus mirabilis (n = 19), Proteus vulgaris (n = 14), Morganella morganii (n = 7), Salmonella More Details enteritidis (n = 23), Salmonella paratyphi A (n = 15), S. paratyphi B (n = 11), Shigella dysenteriae (n = 10), Shigella flexneri (n = 3), Pseudomonas aeruginosa (n = 29), Acinetobacter baumanii (n = 11) and Acinetobacter lwoffii (n = 7) [Table 1].
Beta-lactamase production was demonstrated in 34% (119/350) isolates by the Vitek 2 and phenotypic methods. Both the techniques detected ESBL, AmpC βL, MBLs, ESBL with AmpC, ESBL with MBL and AmpC with MBL in 18.29%, 6%, 1.14%, 2.86%, 3.42%, and 1.43% respectively. Inhibitor resistant TEM (IRT) was observed in 0.86% by the Vitek 2 [Table 1].
Antimicrobial susceptibility testing by the Vitek 2
The resistant levels of ampicillin, cefuroxime, ceftriaxone and cefepime to βL producers were 100%, 89.9%, 86.6 and 73% respectively, but lowered to a range of 44-56% with βL inhibitors (amoxicillin/clavulanate, cefoperazone/sulbactam and pipercillin/tazobactam). Regarding the non-beta lactams, the highest resistant pattern was observed with trimethoprim/sulfamethoxazole (75.6%) followed by nalidixic acid (70.6), ciprofloxacin (66.4%), nitrofurantion (38.6%), and gentamicin (37.8%). Least resistant was demonstrated with colistin (3.4%), tigecycline (5.9%), meropenem (10%), ertapenem or imipenem (12.6%) and amikacin (13.4%) and [Table 2].
A comparison of the activities of the different antibiotics between the βL and non-βL producing isolates, as determined by the Vitek 2 system, is set out in [Table 2]. Among all the tested antimicrobials, the resistance was significantly higher in βL producers than in non-βL producing isolates except for amikacin (P = 0.6), gentamicin (P = 0.4), nitrofurantoin (P = 0.06) and trimethoprim/sulfamethoxazole (P = 0.5).
The present study demonstrated the occurrence of different types of βLs such as ESBLs, AmpC enzymes, MBLs and co-resistant phenotypes (ESBL with AmpC, ESBL with MBL and AmpC with MBL) in our hospital. Three isolates of IRT βLs were detected by the Vitek 2 but final confirmation required molecular techniques, comprising polymerase chain reaction of the coding gene and subsequent sequencing of the product, which could not be carried out in our institute due to lack of infrastructure of genotyping. It was demonstrated that the automated system was able to detect 86.7% of IRT-producing E. coli.
Among the four beta-lactams (cefepime, ceftriaxone, cefuroxime and ampicillin), cefepime was most active against both βL and non-βL producing isolates, although the activity in the latter group was somewhat lower. However, its use is not recommended for ESBLs because of the variable hydrolytic capacity of the enzymes against this antibiotic and failures in the past, despite in vitro activity. 
As far as beta-lactams/βL inhibitors are concerned, pipercillin/tazobactam was the best combination followed by cefoperazone/sulbactam and amoxicillin/clavulanate. Similar finding was observed by Higashitani et al. 
Regarding the aminoglycosides, amikacin was more sensitive than gentamicin against both βL and non-βL producing isolates. Tsai et al. reported that the in vitro activity of amikacin (resistance rate of 2.9%) and isepamicin was superior to other tested aminoglycosides (gentamicin and tobramycin) against Gram-negative bacteria isolates.  Livermore et al. observed resistance of 25% and 40% for amikacin and gentamicin respectively against carbapenem resistant Enterobacteriaceae isolates. 
The fluoroquinolones assayed showed less activity against βL producers, probably because of widespread use of these antimicrobials against infections caused by βL producing microorganisms that do not respond to treatment with beta-lactams in recent years. Clonal spreading might be another reason for the emergence of fluoroquinolone resistance at a rapid rate.  De Francesco et al. and Tankhiwale et al. reported ciprofloxacin resistance rate of 89.9% and 69% respectively for ESBLs. ,
Interestingly, nitrofurantoin showed resistance of 38.6% and 28.9% against βL and non-βL producing isolates respectively. This antibiotic has been used for more than five decades for the treatment or prevention of uncomplicated cystitis and still remains active against most of the uropathogens. The emergence of resistant variants from initially susceptible pathogens has been rare, despite many years of clinical use. The absorption of oral nitrofurantoin is 40-50% and it is enhanced when it is taken with food.  Its serum concentrations are low to undetectable, and urine concentrations are 50-250 mg/ml.  The occurrence of resistance to nitrofurantoin for ESBLs in recent surveys in the USA and Canada was 1.1% among 1142 isolates of E. coli from out-patient urinary isolates. 
As far as susceptibility to trimethoprim/sulfamethoxazole (cotrimoxazole) was concerned, the isolates were divided evenly, implying that this antibiotic was not really useful in these cases unless an antibiogram was performed. Our study demonstrated a resistant rate of 75.6% to this antibiotic that was comparable to that of Pai et al. (71.38%). 
Among the carbapenems, meropenem proved to be active against 90% of the βL producers in comparison to imipemen (88%) and ertapenem (88%). However, they remained sensitive to all the non-βL producing isolates. According to these results, meropenem, and consequently the carbapenems in general, would seem to be the best therapeutical choice for hospital use particularly in combating infections caused by ESBL and AmpC producing microorganisms. Thomson and Moland studied the activity of meropenem, cefepime, and piperacillin/tazobactam in vitro and observed that, at least in those clinical cases where there was a high bacterial inoculate such as endocarditis, meningitis, septic arthritis, osteomyelitis, and abscesses, the best choice of treatment was meropenem.  However, the clinical utility of these antimicrobials is under threat with the emergence of carbapenemases, including MBLs.
In our study, colistin followed by tigecycline proved to be most effective antimicrobials against βL producers. Tigecycline may be a therapeutic alternative to carbapenems in some infections caused by ESBL and AmpC producing Enterobacteriaceae, many of which are also multiresistant to quinolones, aminoglycosides and classical tetracyclines.  However, limitations are that many infections with ESBL producers occur in the urinary tract, whereas tigecycline has largely biliary excretion with low urinary recovery and few isolates of ESBL-positive E. coli, Klebsiella spp. and Enterobacter spp. do have intermediate resistance to the compound.  Chandran et al. suggested that they are the only therapeutic options available currently for treating NDM-1 producing Gram-negative infections. 
Colistin, tigecycline and meropenem proved to be the most suitable antibiotics against infections by βLs producing microorganisms because βLs offer considerable resistance to other antibiotic groups. In our evaluation, imipenem or ertapenem, amikacin and nitrofurantoin also proved to be suitable alternatives.
Source of Support: None, Conflict of Interest: None
[Table 1], [Table 2]