What is DISCUSSION ?

The critically ill patients such as heart patients are at higher risk of acquiring infection due to multiple reasons including disruption of barrier to infection by invasive instrumentational procedures such as endotracheal intubation and tracheostomy, urinary bladder catheterization, central venous catheterization etc. (Shannon 2005; Shaikh et al 2008). Due to the extensive use of broad spectrum antibiotics in ICU patients and production of antibiotic hydrolyzing enzymes, bacterial isolates are resistant to many antibiotics (Fridkin et al 2001; Kollef et al 2001). The prevention of such infections demands through knowledge of infection rates and source, type and nature of invading microorganisms along with risk factors associated with infection (Weinstein 1991). In the context of under developed countries like Nepal, there are hardly any such studies carried out to identify the etiological agent of infection and their antibiotic resistance pattern among the patients. So, this study was carried out to determine the prevalence of multidrug resistant bacterial isolates capable of producing beta lactamase enzymes from patients visiting Sahid Gangalal National Heart Centre, Bansbari. Intensive Care Unit (ICU) is a specialized area of a hospital where critically ill patients requiring intensive care are treated. Of all other infections, the rate of infection is highest in ICU patients (Tennant et al 2005).

Of the 357 samples consisting of various clinical specimens (urine, blood, catheter tips, sputum and pus), only 98 Gram negative (27.455%) showed significant growth. Of the total isolates, 76 (77.55%) were found to be multidrug resistance and 82(83.6%) were from those admitted in ICU. A similar study conducted by Poudyal (2010) in National Public Health Laboratory showed 19.61% growth and 61.27% of MDR among the isolates and in a study conducted by Panta (2013) in National Kidney Centre showed 19.92% growth and 85.83% of MDR. Baral (2008) showed the low culture positivity of 22.35% among various samples and multidrug resistance i.e. 41.07% among various clinical isolates. In a study by Bomjan (2005) a high proportion (60%) of multidrug resistance pattern among the urinary and sputum isolates was reported. Pokhrel (2005) in a study of urinary and sputum isolates in TUTH showed 47.57% and 60.40% of the sputum and urinary isolates respectively were multidrug resistance. A similar study carried out by Shrestha (2009) conducted at Kathmandu Model Hospital showed 55.8% prevalence of MDR. Similar results showing greater prevalence of Gram negative bacteria in the ICU patients has also been reported in many other studies (Hassanzadeh et al 2009;  Vincent et al 2009 ;Tennant et al 2005).

Most of the isolates in this study were especially from urine followed by the catheter tips (34/98=34.7%) which is in accordance to several other studies (Reshedko et al 2007). Most of the samples acquired for culture were of age group more than 50 years, which is in accordance with the studies conducted by Falagas (2000), Naqvi and Collins (2006). However, the association between age group and culture positivity was statistically insignificant.

Only the Gram negative isolates were further processed for AST and other microbiological testings. This is because most of the predominant pathogens isolated from heart patients are of Gram negative class. Of the 98 Gram negative isolates, E. coli was the predominant pathogen adding up to 33.67% followed by P. aeruginosa, 29.59% of the total isolates. High numbers of Gram negative isolates were reported by the investigations conducted by Baral (2008), Bomjan (2005), Falhal et al (2012), Manandhar (1996), Panta (2013) ,Poudyal (2010) and Puri (2006). These results were in harmony with the results obtained in similar studies by Baral (2008), Dhakal (1999), Farrell et al (2003), Gales et al (2002), Kahlmeter (2000), Mathai et al (2001) and Poudyal (2010). However, lesser number of P. aeruginosa  was isolated. Similarly, P. aeruginosa, E. coli and Acinetobacter spp. were other major aetiological agents isolated from  the catheter tips samples. The reasons behind such results may due to the nosocomial transmission, fastidious bacteria not easily grown on routine culture media, delay in transport and several other factors.

Of the 33 E. coli isolates, 26 (78.78%) were multidrug resistant.75% K. pneumoniae, 82.7% P. aeruginosa, 66.6% C. freundii, 60%  Enterobacter spp., 75% Acinetobacter spp.,and 100% Salmonella Typhi were MDR. These results reassembled the outcomes of previous studies by Baral (2008), Pokhrel (2005), Bomjan (2005), Poudyal (2010) and Panta (2013).

Colonization is a necessary prerequisite for subsequent infection caused by Multidrug resistant Gram negative bacteria (MDRGN. Translocation of MDRGN across the intestinal wall into the blood stream and fecal contamination of vascular devices lead to infections (Tancrede et al 1985). Thus, patients who are colonized with MDR Gram negative bacteria are at greater risk for subsequently developing an infection with these bacteria. In one study, 15% of hospitalized patients who were colonized with MDR Gram negative bacteria developed a bacteremia caused by the same colonizing strain of MDR Gram negative bacteria (Ben-Ami et al 2006). The co-resistance to multiple antimicrobials among MDR Gram negative bacteria severely limits the therapeutic options that are available to physicians for treating infections that are caused by MDR Gram negative bacteria.

Transferable resistance has been identified for some antibiotic groups as β-lactams, aminoglycosides, macrolides, sulphonamides, tetracyclins, chloramphenicol, etc. However the production of plasmid or chromosomal encoded β-lactamase enzymes is the most common mechanism of resistance in Gram negative bacteria causing clinical significant infection (Bush 1995). Endogenous acquisition, as opposed to patient-to-patient spread, is the predominant mechanism of acquisition. Residence in a long-term care facility and antibiotic exposure may be important factors promoting the spread of multidrug-resistant gram-negative bacteria among this patient population (Pop-vicas  et al 2008).

The high level of drug resistance seen among E. coli is due mediated by β-lactamases, which hydrolyze the β-lactam ring inactivating the antibiotic, The classical TEM-1, TEM-2, and SHV-1 enzymes are the predominant plasmid-mediated β-lactamases of Gram-negative rods (Livermore, 1995). Mutations at the target site i.e. gyrA, which is a gyrase subunit gene, and parC, which encodes a topoisomerase subunit, confer resistance to fluoroquinolones (Ozeki et al 1997). In addition to this mechanism, there are more than seven efflux systems in E. coli that can export structurally unrelated antibiotics; these multidrug resistance efflux pump (MDR pump) systems contribute to intrinsic resistance for toxic compounds such as antibiotics, antiseptics, detergents, and dyes (Sulavik et al 2001).

Similarly higher level of drug resistance seen among K. pneumonia and Acinetobacter spp. is mediated by the production of different kind of β-lactamases primarily ESBLs, AmpCs and MBLs. The fact that the carriage of resistance trait for quinolones and aminoglycoside in the plasmid along with the gene for β-lactamases have had a great impact on the drug resistance character shown by these pathogenic bacteria (Lee et al 2003; Picao et al 2008; Thomson et al 2000 and Walsh et al 2005). Moreover, various clinical isolates show alteration of nonspecific porins associated with the presence of active drug efflux in these bacteria; both processes maintain a very low intracellular concentration of drugs and contribute to a high resistance level for structurally unrelated molecules including β-lactam antibiotics, quinolones, tetracyclines, and chloramphenicol (Martinez-Martinez et al 2002).

Imipenem with susceptibility of 83.6% was found to be the most effective drug against Gram negative isolates. However, it is the alternative therapeutic agent in absence of other first line drugs. Among others, meropenem, nitrofurantoin, amikacin and gentamicin were found to be more effective with 79.5%, 66.3% , 63.2%  and 61.2% susceptibility respectively. Amoxycillin, ceftazidime and nalidixic acid with the sensitivity of 10.2%, 19.3% and42.8% respectively were the least effective drugs. As these antibiotics are the first line drugs which are easily hydrolysed by the bacterial enzymes and offer less in the treatment of Gram-negative bacterial infections. In concurrence to our findings, imipenem was found to be the most effective drug against Gram negative isolates in the study of Baral (2008), Oteo et al (2001), Panta (2013) and Puri (2006). However, Poudyal (2010) found that meropenem was more effective than imipenem. Meropenem is more stable than imipenem to  kidney enzyme dehydropeptidase and can be administered without cialistin hence, may be better therapeutic option than imipenem (Jones et al 2008).

First described in 1983, ESBLs have contributed to dramatic increase in resistance to beta-lactamase among Gram negative bacteria in recent years. A total of 64 multi-drug resistance bacteria were screened for ESBL production using two of the CLSI recommended screening agents viz. ceftazidime and cefotaxime. The lowest sensitivity (89.13%) was observed with ceftazidime, when the screen positive isolates were subjected to ESBL confirmation using inhibitor potentiated disk diffusion (IPDD) test. Cefotaxime was comparatively more sensitive towards ESBL screening (97.82%). Poudyal (2010) also found that cefotaxime is more effective in screening ESBL than ceftazidime and the percentage of sensitivity is in harmony with our result.However, Panta (2013) found lowest sensitivity with cefotaxime. Katz et al (2004) subjected 115 isolates of E. coli and 157 isolates of Klebsiella spp. for screening using cefotaxime, ceftazidime, and cefpodoxime disks. The sensitivity of screening criteria ranged between 98.6% for cefotaxime and 92.8% for ceftazidime, and the specificity ranged between 100% for cefotaxime and cefpodoxime and 99.0% for ceftazidime. Similar results were obtained in the study of Jain et al (2007).

For the confirmation of ESBL production, 2 different combination disks were used in our study, The inhibitor potentiated disk diffusion test (IPDDT) identified 46 suspected isolates as confirmed ESBL producers. The Ceftazidime  and Ceftazidime-clavulanate combination disk correctly identified only 41 isolates as confirmed ESBL producers.However, cefotaxime-clavulanate combination disk correctly identified all the ESBL producing isolates. Similar pattern of results were obtained in the study carried out by (Jain et al 2007;  Poudyal  2010).

In our study, of the 15 isolates sensitive to ceftazidime (Screen negatives) in disk diffusion test, 5 isolates were found to be ESBL producers when tested with other IPDDT suggesting the possible presence of CTX-M type ESBL, however, due to lack of genetic characterization of the enzyme, it could not be confirmed. CTX-M-type enzymes were reported in Germany and Argentina in 1989, and so far, more than 67 CTX-M-type β-lactamases have been identified, mostly in E. coli, K. pneumoniae, and S. enterica serovar Typhimurium isolates (Gonullu et al 2008). Diagnostic laboratories may fail to identify CTX-M–positive isolates as ESBL producers if ceftazidime resistance is used as the sole screening criterion since CTX-M producing isolates have typical propensity towards cefotaxime, however, are susceptible to ceftazidime in vitro. CTX-M extended spectrum β-lactamases (ESBLs) differ from those derived from TEM and SHV enzymes by their preferential hydrolysis of cefotaxime and ceftriaxone compared with ceftazidime (Lewis et al 2007).

Of the 64 bacterial isolates consisting of 7 genera tested for the ESBL production, 44(68.7%) isolates tested positive for ESBL production. The majority consisted of E. coli i.e. 18/44 (86.96%) followed by P. aeruginosa 10/44 (5.8%). Two isolates each of K. oxytoca and K. pneumoniae  , 3 of Acinetobacter spp.,3 of C. freundii,5 of Enterobacter spp. and 1 of S. Typhi showed ESBL production. Similar pattern of results were seen in the study carried out by Poudyal (2010) who showed the presence of 62.72% ESBL producers out of 110 MDR isolates,  Bomjan (2005) who found the presence of 28.3% ESBL producers among various clinical isolates and Sharma (2004) who found 8% K. pneumoniae, 12.5% E. coli, 12.5% C. freundii, 25% A. calcoaceticus and 5% P. aeruginosa as ESBL-producing strains. In the  study performed by Paudel (2013), of 267 Enterobacteriaceae 72 (27%) were ESBL producers. From the data of SMART program in the Asia-Pacific region, of 3,004 Gram-negative bacilli collected from intra-abdominal infections during 2007, 42.2% and 35.8% of E. coli and Klebsiella spp. respectively were ESBL positive. Moreover, ESBL rates in India for E. coli, K. pneumoniae, and K. oxytoca were 79.0%, 69.4%, and 100%, respectively. ESBL-positive E. coli rates were also relatively high in China (55.0%) and Thailand (50.8%) (Hawser et al 2009).

In the present study all the ESBL producers were resistant to Amoxycillin. This is due to the enzyme beta-lactamase that inactivates the antibiotics and renders them ineffective. Similar resistance pattern was observed in the study carried out by Manandhar (2006), Pokhrel (2006) and Poudyal (2010). Of all the ESBL producers all were resistant to five or more than five of the most commonly used antibiotics. In a similar type of study conducted by Poudyal (2010), ESBL producers were significantly resistant to commonly used antibiotics than non-ESBL producers and multidrug resistance was significantly higher in ESBL positive isolates than non-ESBL isolates.

In those organisms that typically have inducible AmpC chromosomal enzyme,it may be induced by clavulanate and attack the indicator cephalosporin, thus masking any synergy arising from Esbl production.So, ESBLs are rarer in Enterobacteriaceae other than E. coli and Klebsiella spp. and are more difficult to detect. In case of C. freundii, M. morganii, Serratia spp., Providencia spp. and Enterobacter spp. resistance to oxyiminocephalosporin is due to mutational hyperproduction of chromosomal AmpC production rather than ESBL. AmpC inducible species segregate ‘depressed’ mutants which copiously produce AmpC enzyme without induction and are resistant to almost all penicillins and cephalosporins (Livermore et al 2001; Freeman et al 2009).

In our study, 61 MDR strains were resistant to ceftazidime and with reduced carbapenem susceptibility were subjected for the detection of MBL production, of which 35 (57.3%) tested positive for MBL production using Imipenem-EDTA combined disk test. Of the 35 MBL producers, 14 (40%) isolates were P. aeruginosa, followed by 9 (25.71%) of E. coli isolates. Five isolates of Acinetobacter spp., 2 isolates each of C. freundii, Enterobacter spp. and K. oxytoca and 1 isolate of K.  pneumoniae were MBL producer. In a study conducted in France in 241 clinical strains of IPM-nonsusceptible Ps. aeruginosa isolated from 2002 to 2004, 110/241 (46%) were MBL positive using phenotypic methods while 107/241 (45%) were PCR positive for MBL genes: 103/241 (43%) for bla_VIM and 4/241 (2%) for bla_IMP (Pitout et al., 2005)_. Yan and colleagues reported an outbreak of K. pneumoniae isolates carrying blaIMP-8 and showed that 88% (35/40) were susceptible to carbapenems. In a similar study conducted by Baral (2008), of the 117 MDR isolates, 33 (28.2%) isolates were MBL producers among them 25 (75.75%) were E. coli isolates and 3 (9.1%) C. freundii isolates. In a similar study conducted by Baral (2008), of the 117 MDR isolates, 33 (28.2%) isolates were MBL producers among them 25 (75.75%) E. coli isolates and 3 (9.1%) C. freundii isolates.In a similar study conducted by Poudyal (2010), of the 59 MDR isolates, 25 (42.37%) isolates were MBL producers among them 12 (48%) were E.coli isolates followed by 4 (16%)  each of K.  pneumoniae and Acinetobacter spp. Zavascki et al (2005) has reported the presence of 77.1% MBL producers in Southern Brazil upon testing 35 isolates of Carbapenem resistant Ps. aeruginosa (CRPA), among which 27 were MBL positive.

In a study conducted in Italy, of 14,812 consecutive non replicate clinical isolates (12,245 Enterobacteriaceae isolates and 2,567 Gram-negative nonfermenters) screened for reduced carbapenem susceptibility, 30 isolates ( 28 P. aeruginosa isolates, 1 P. putida isolate, and 1 E. cloacae isolate) carried acquired MBL determinants (Rossolini et al 2008). MBL producers were detected in 10 of 12 cities, with a predominance of VIM-type enzymes over IMP-type enzymes (4:1). Since there are no standardized phenotypic methods available for the detection of MBL, several tests have been employed to detect the MBL production depending upon whether the gene is carried by P. aeruginosa or a member of Enterobacteriaceae and taking advantage of chelating agents, EDTA and thiol based compounds, to inhibit MBL activity (Walsh et al 2005). Franklin et al (2006) found a sensitivity and specificity of 100% and 98% using an Imipenem(10 µg) and Imipenem-EDTA (292µg) combined disk test for MBL detection respectively. the ability of the bacteria to produce MBL was compared with their ability to hydrolyse carbapenems which was found statistically significant p<0.05.

Due to emergence of acquired carbapenemases i.e MBL, the clinical utility of carbapenems is under threat. Since the early 1990s, new MBL encoding genes have been reported all over the world in clinically important pathogens, such as Pseudomonas spp., Acinetobacter spp. and members of Enterobacteriaceae family. The emergence of MBL-encoding genes is worrisome since they are usually carried by mobile genetic cassettes inserted in integrons with great ability to spread. Moreover, increased mortality rates have been documented for patients infected with MBL-producing Ps. aeruginosa, especially due to inadequate emperical therapy. Therefore, early detection of MBL-producing organism is crucial to start appropriate antimicrobial therapy (Franklin et al 2006). The phenotypic appearance of MBL-carrying organisms varies depending on the bacterial host, with increasing reports of carbapenem-susceptible isolates, primarily Enterobacteriaceae such as Klebsiella spp. and E. coli. Such organisms often carry hidden MBL genes, whereby the microbiologist and the clinician remain unaware of their presence within an institution. Such a scenario creates the potential for untoward clinical and infection control consequences and is by no means unique to MBLs (Franklin et al 2006).