Phenotypic characterization of carbapenem non-susceptible gram-negative bacilli isolated from clinical specimens

Background Multidrug resistant, extremely drug-resistant, pan-drug resistant, carbapenem-resistant, and carbapenemase-producing gram-negative bacteria are becoming more common in health care settings and are posing a growing threat to public health. Objective The study was aimed to detect and phenotypically characterize carbapenem no- susceptible gram-negative bacilli at the Ethiopian Public Health Institute. Materials and methods A prospective cross-sectional study was conducted from June 30, 2019, to May 30, 2020, at the national reference laboratory of the Ethiopian Public Health Institute. Clinical samples were collected, inoculated, and incubated for each sample in accordance with standard protocol. Antimicrobial susceptibility testing was conducted using Kirby-Bauer disk diffusion method. Identification was done using the traditional biochemical method. Multidrug-resistant and extensively drug-resistant isolates were classified using a standardized definition established by the European Centre for Disease Prevention and Control and the United States Centers for Disease Prevention and Control. Gram-negative organisms with reduced susceptibility to carbapenem antibiotics were considered candidate carbapenemase producers and subjected to modified carbapenem inactivation and simplified carbapenem inactivation methods. Meropenem with EDTA was used to differentiate metallo-β-lactamase (MBL) from serine carbapenemase. Meropenem (MRP)/meropenem + phenylboronic acid (MBO) were used to differentiate Klebsiella pneumoniae carbapenemase (KPC) from other serine carbapenemase producing gram-negative organisms. Results A total of 1,337 clinical specimens were analyzed, of which 429 gram-negative bacterial isolates were recovered. Out of 429 isolates, 319, 74, and 36 were Enterobacterales, Acinetobacter species, and Pseudomonas aeruginosa respectively. In our study, the prevalence of multidrug-resistant, extensively drug-resistant, carbapenemase-producing, and carbapenem nonsusceptible gram-negative bacilli were 45.2%, 7.7%, 5.4%, and 15.4% respectively. Out of 429 isolates, 66 demonstrated reduced susceptibility to the antibiotics meropenem and imipenem. These isolates were tested for carbapenemase production of which 34.8% (23/66) were carbapenemase producers. Out of 23 carbapenemase positive gram-negative bacteria, ten (10) and thirteen (13) were metallo-beta-lactamase and serine carbapenemase respectively. Three of 13 serine carbapenemase positive organisms were Klebsiella pneumoniae carbapenemase. Conclusion This study revealed an alarming level of antimicrobial resistance (AMR), with a high prevalence of multidrug-resistant (MDR) and extremely drug-resistant, carbapenemase-producing gram-negative bacteria, particularly among intensive care unit patients at the health facility level. These findings point to a scenario in which clinical management of infected patients becomes increasingly difficult and necessitates the use of “last-resort” antimicrobials likely exacerbating the magnitude of the global AMR crisis. This mandates robust AMR monitoring and an infection prevention and control program.


Introduction
The discovery of the antimicrobial agent is a fundamental milestone in the history of medicine and has saved millions of lives [1]. Antimicrobials were first used to treat infections in the 1940s [2]. Shortly after the discovery of antimicrobials, antimicrobial resistance emerged and with the recent increase in AMR, poses a serious threat to global public health [2].
The extensive use of antimicrobials for treating humans and animal infections along with globalization and international travel has led to the rapid spread of resistant strains [3]. The increasing incidence of healthcare-associated infections due to multidrug-resistant (MDR), extremely drug-resistant (XDR), and carbapenemase-producing gram-negative bacilli (GNB) has been widely reported [4][5][6][7].
The emergence and spread of multidrug-resistant gram-negative organisms (MDRO) pose serious threats to medical services and patient outcomes. Infections caused by carbapenemaseproducing and carbapenem resistant Enterobacterales, Acinetobacter baumannii, Pseudomonas aeruginosa result in increased patient morbidity and mortality. This can result in significant additional health-care costs for patient management as well as outbreak control [8][9][10][11].
Multidrug resistance has increased globally and is a major public health threat. Recent investigations reported the emergence of multidrug-resistant bacterial pathogens from different origins including humans, poultry, cattle, and fish that increase the need for routine application of antimicrobial susceptibility testing (AST) to select the antibiotic of choice as well as screening of emerging MDR strains [17][18][19][20][21][22][23][24].
International subject matter experts came together through a joint initiative by the European Centre for Disease Prevention and Control (ECDC) and the United States Centers for Disease Control and Prevention (CDC), to create a standardized international definition with which to describe acquired resistance profiles [25].
MDR is defined as acquired nonsusceptibility to at least one agent in three or more antimicrobial categories, XDR as nonsusceptibility to at least one agent in all but two or fewer antimicrobial categories (i.e., bacterial isolates remain susceptible to only one or two categories), and PDR (Pandrug-resistant) as nonsusceptibility to all agents in all antimicrobial categories. To apply these definitions, bacterial isolates should be tested against all or nearly all of the antimicrobial agents within the antimicrobial categories and selective reporting and suppression of results should be avoided [25].
There is a considerable knowledge gap regarding risk factors associated with the occurrence of MDR bacteremia [28]. Identifying risk factors for acquiring gram-negative bacteremia could potentially help patient care and management [28].
Despite the increasing global burden of multidrug resistance and carbapenemase-producing gram-negative bacilli, data on multidrug resistance and carbapenemase-producing gramnegative bacilli in Ethiopia is scarce. As a result, the objective of this study was to determine the prevalence of MDR, XDR, carbapenem non-susceptible and carbapenemase-producing gram-negative bacilli from various clinical specimens and to phenotypically characterize carbapenem non-susceptible isolates.
according to standard protocols related to each sample [29]. Identification was done using the conventional biochemical method [29]. Gram staining, colony characterization, and biochemical tests were conducted as part of the identification process. AST was done by the Kirby Baur disk diffusion method as per CLSI M100 2020. All frequently isolated Enterobacterales, Acinetobacter species, and P. aeruginosa recovered from the various clinical specimens during the study period were included. Necessary variables such as socio-demographics (age and sex), specimen type, referring health facilities, patient location, and previous antibiotic exposure from the test request form were entered onto pre-configured WHONET software version 2019.

Bacterial isolation, identification and antimicrobial susceptibility testing
Bacterial isolation and identification. The specimens were inoculated onto appropriate culture media, incubated at appropriate temperature and time following standard procedure [29]. The growth was inspected to identify the bacteria. Initial identification of bacteria was done based on Gram reaction and colonial morphology. Gram-negative rods were identified by performing a series of traditional biochemical enzymatic and carbohydrate fermentative tests on triple sugar iron agar (Liofilchem, Roseto degli Abruzzi, Italy), oxidase strips (Liofilchem, Roseto degli Abruzzi, Italy), Simon's citrate agar (Liofilchem, Roseto degli Abruzzi, Italy), and lysine iron agar (Liofilchem, Roseto degli Abruzzi, Italy). Indole production and motility were obtained using sulfide-indole-motility (SIM) medium (Liofilchem, Roseto degli Abruzzi, Italy). Urease production was obtained using a urea agar base supplemented with 40% urea solution (Oxoid Ltd., Basingstoke, Hampshire, England).
Antimicrobial susceptibility testing. The Kirby Bauer disk diffusion method was used with Muller Hinton agar (Oxoid Ltd. Basingstoke, Hampshire, England) to determine antimicrobial susceptibility patterns of the isolates and CLSI M100 2020 was used to interpret the results [30].

Detection of carbapenemase
The Clinical and Laboratory Standards Institute CLSI (2010) introduced the modified Hodge test for carbapenemase detection, but this method can only be used for the accurate detection of KPC-type carbapenemase in Enterobacterales [31]. CLSI (2012) recommended the Carba NP test method for the detection of carbapenemase in gram-negative bacilli; however, the preparation of the reagents required for this test is complicated and the solutions cannot be stored for extended periods, limiting its clinical application [32].
In 2015, the carbapenem inactivation method (CIM) which is easy to operate and highly sensitive for the detection of carbapenemase was introduced [33]. In 2017, based on the CIM method, CLSI recommended the modified carbapenem inactivation method (mCIM), However, it is a relatively complex method and can only be used to detect carbapenemase in Enterobacterales and P. aeruginosa [34]. In 2018, based on the mCIM, a simplified carbapenem inactivation method (sCIM) was designed for accurate detection of carbapenemase in gramnegative bacilli [35].
Modified carbapenem inactivation method. In the mCIM, a 1 mL loopful of Enterobacterales or a 10 mL loopful of P. aeruginosa from blood agar (Oxoid Ltd., Basingstoke, Hampshire, England) plates was emulsified in 2 mL trypticase soy broth (TSB) (Oxoid, Ltd., Basingstoke, Hampshire, England). A meropenem (10μl) disk (Oxoid Ltd., Basingstoke, Hampshire, England) was then immersed in the suspension and incubated for a minimum of 4 h at 35˚C. A 0.5 McFarland suspension of E. coli ATCC 25922 was prepared in saline using the direct colony suspension method. A Mueller Hinton agar (MHA) (Oxoid Ltd., Basingstoke, Hampshire, England) plate was inoculated with E. coli ATCC 25922 using the routine disk diffusion procedure. The meropenem disk was removed from the TSB and placed on a MHA plate previously inoculated with the E. coli ATCC 25922 indicator strain. Plates were incubated at 35˚C in ambient air for 18-24 h. An inhibition zone diameter of 6-15 mm or colonies within a 16-18 mm zone was considered to be a positive result, and a zone of inhibition of �19 mm was considered to be a negative result [30].
Simplified carbapenem inactivation method. The sCIM is based on the mCIM with an improved procedure. Instead of incubating the antimicrobial disk in the organism culture media for 4 hours as in the mCIM, the organism to be tested was smeared directly onto an antimicrobial disk in the sCIM. To perform the sCIM for Acinetobacter species, a 0.5 McFarland standard suspension (using direct colony suspension method) of E. coli ATCC 25922 was diluted 1:10 in saline and inoculated onto the MHA plate, following the routine disk diffusion procedure. Plates were allowed to dry for 3-10 min [35].
Subsequently, 1-3 colonies of the test organisms grown overnight on blood agar were smeared onto one side of an imipenem disk (10μg); immediately afterward, the side of the disk having bacteria was placed on the MHA plate previously inoculated with E. coli ATCC 25922. The imipenem disk placed on an MHA plate was used as the control [35].
All plates were incubated at 35˚C for 16-18 hours in ambient air. Bacterial strains that produced carbapenemase hydrolyze imipenem; hence the susceptible indicator strain grows unchecked. If the zone of inhibition around the disk gave a diameter of 6-20 mm, or the satellite growth of colonies of E. coli ATCC 25922 around the disk a zone diameter of � 22 mm, the result was considered carbapenemase positive; a zone of inhibition � 26 mm was considered to be a negative result; a zone of inhibition of 23-25 mm was considered to be a carbapenemase indeterminate result [35].
Differentiation of metallo-β-lactamase from serine carbapenemases. The modified carbapenem inactivation method positive Enterobacterales (formerly Enterobacteriaceae) and P. aeruginosa, and the Simplified Carbapenem Inactivation Method positive Acinetobacter species detected were further screened for Class B metallo-carbapenemase (MBLs) which are characterized by inhibition with metal chelators like EDTA. Meropenem disks containing EDTA (Liofilchem, Roseto degli Abruzzi, Italy) were used to differentiate metallo-β-lactamase and serine carbapenemase. A � 5-mm increase in zone diameter for eCIM vs. zone diameter for mCIM was considered MBL positive. A �4mm increase in zone diameter for the eCIM vs zone diameter of mCIM was considered MBL negative. Carbapenemase positive, metallo carbapenemase negative, gram-negative bacilli were considered serine carbapenemase producers [30].
Well isolated colonies from an overnight blood agar plate were suspended in saline to achieve a 0.5 McFarland standard turbidity (1.0 McFarland if mucoid) to obtain a confluent lawn of growth after incubation. The strip was applied to the agar surface with the scale facing upwards and the code of the strip to the outside of the plate. The agar plates were incubated in an inverted position at 35 ± 2˚C for 16-20 hours in an ambient atmosphere. The incubation time was extended to 48 hours to capture any slow-growing gram-negative non-fermenters [36].
Result interpretation. The ratio of MRP/MBO of �8 or �3 log2 dilutions was interpreted as a KPC producer. The phantom zone or deformation of the ellipse was interpreted as positive for KPC regardless of the MRP/MBO ratio [36].
Quality control recommendations. K. pneumoniae ATCC1 BAA-1705 (intrinsic KPC production) and E. coli ATCC1 25922 and were used as positive and negative controls respectively to check the reactivity of KPC strips [36].

Reagent quality assurance
The quality of culture media, antimicrobial disks, and gradient strips were checked as per CLSI and EUCAST guidelines, laboratory SOPs, and the manufacturer's instructions as applicable.

Data analysis and interpretation
The WHONET 2019 version was used to enter, clean, and analyze the data. The risk factors for MDR gram negative bacilli acquisition were examined by exporting WHONET data to SPSS statistics version 23. Tables and figures were used to present the results. Chi-square and univariate analysis were used to determine the association between multidrug-resistant gram-negative bacilli and different risk factors. P-values less than 0.05 were considered statistically significant.

Ethical considerations
The study was conducted after ethical clearance was obtained from the Department Research and Ethical Review Committee of the Department of Medical Laboratory Sciences, College of Health Sciences, Addis Ababa University. Official permission from the Ethiopian Public Health Institute was obtained. All results were kept confidential; the patient's name and other personal identifiers were encrypted and the sample identifier automatically generated by the laboratory information system (LIS-Polytech) was used.

PLOS ONE
Phenotypic characterization of carbapenem non susceptible gram-negative bacilli The prevalence of multi-drug resistant and extremely drug resistant, and carbapenemase-producing GNB. Of 194 MDR GNB isolates, 45% were isolated from patients admitted to the intensive care unit, 83.4% were isolated from patients previously exposed to different antimicrobial agents, 28% were isolated from patients under mechanical ventilation and/or urinary catheterization, and 34.7% were isolated from patients with hospital-acquired (HAI) pneumonia (Table 3).  (Table 3).
K. ozaenae showed 100% resistance to first, second, and third generation cephalosporins as well as trimethoprim-sulfamethoxazole and Enterobacter cloacae demonstrated 100% resistance to ceftriaxone. The overall resistance profile of Enterobacterales to extended-spectrum (3 rd Table 5). The overall prevalence of Enterobacterales resistance to gentamycin, tobramycin, and amikacin was 41.9%, 37.5%, and 6.1% respectively. The overall prevalence of meropenem and imipenem resistance among the Enterobacterales was 4.2% and 6.5% respectively.
Morganella morganii, P. mirabilis, and P. vulgaris were reorganized into the Morganellaceae family [30]. A total of 26 Morganellaceae isolates were identified during the study period ( Table 2). M. morganii and P. vulgaris are intrinsically resistant to ampicillin, cefazolin, and cefuroxime and were not reported in Table 5. M. morganii, P. mirabilis, and P. vulgaris intrinsically resistant to antimicrobials (

PLOS ONE
Phenotypic characterization of carbapenem non susceptible gram-negative bacilli percentages of meropenem and imipenem resistance were 14% and 14.3% respectively ( Table 5). The antimicrobial susceptibility patterns of the Morganellaceae are summarized in Table 5. The percentage of Morganellaceae resistance among aminoglycosides ranges from 26.3% to 47.5% (Table 5).     (Table 5).

Extensively drug resistant gram-negative bacilli
The most frequently isolated XDR organism was Acinetobacter species 32.4% [24/74] [38]. The difference could be attributed to geographical differences, the number of samples studied, or the types of gram-negative bacteria considered. In the present study, the highest number of XDR organisms were recovered from urine samples 48.48% [16/ 33] and patients admitted to an intensive care unit 63.64% [21/33].
In the present study, the prevalence of XDR gram-negative bacilli was 7.7% [33/429], which is slightly lower than a study at Ethiopian Public Health Institute, Ethiopia by Beyene et al 8.8% [37]. This variation might be due to the investigators analyzing only Enterobacterales and is much lower than the findings from the study at Jimma, Ethiopia by Gashaw et al 41.3% [38], Arsho Advanced Medical Laboratory, Addis Ababa, Ethiopia by Bitew et al 34.3% [39], and the Tertiary Care Hospital, Pakistan by Abbas et al 64% [40]. The variation might be due to geographic location, the technique utilized, XDR definition, types of organism, etc.

Carbapenemase producing gram negative bacilli
In our study, the prevalence of carbapenemase-producing gram-negative bacilli was 5.4%, which is higher than the prevalence of study conducted at the University of Gondar, Ethiopia by Eshetie et al at 2.72% [41] and the Ethiopian Public health Institute by Beyene et al at 2% [37]. However, our result was lower than the result of studies conducted at Tikur Anbessa Specialized Hospital, Ethiopia by Melese et al with 12.12% [42], Three Hospitals in Amhara region, Ethiopia by Moges et al 15.7% [43], Felegehiwot Hospital, Ethiopia by Moges et al 16.2% [44], Sidama, Ethiopia by Alemayehu et al 9% [45], Mulago National Referral Hospital, Uganda by Okoche et al 22.4% [46], and data from laboratories in Spain by Lopez-Hernandez et al 62% [47]. The variation might be due to the method utilized, i.e. the modified carbapenem inactivation method was utilized in our investigation unlike other investigators who used the modified Hodge test, the number of bacterial isolates analyzed, geographic location. However, our study findings are in line with a prospective cross-sectional study conducted at Felegehiwot Hospital, Ethiopia by Alebel et al 5.2% [48]. The dominant carbapenemase-producing gram-negative bacilli were Klebsiella pneumoniae 8.3% [9/109] followed by Acinetobacter species 6.8% [5/74] [44], Lopez-Hernandez Klebsiella pneumoniae 45% [47]. The present study disagrees with the findings by Okoche et al, which showed E. coli as the highest carbapenemase-producing organisms [46], The discrepancy could be explained by the fact that they utilized boronic acid-based inhibition, a modified Hodge test and EDTA double combination disk test, the number of samples analyzed, and a different geographical location.
The highest number of carbapenemase producing isolates were recovered from urine samples 13/23 (56.5%), which strongly disagrees with the results of study conducted by Moges et al in which the highest number of carbapenemase producing isolates was recovered from blood samples 22/24 (91.6%) [43].  [37]. The observed variation could be attributable to the study period and proportion of bacterial isolates, Alebel et al in which Acinetobacter species (100%), P. aeruginosa (100%), Citrobacter species (100%), and Enterobacter cloacae (100%) were the dominant MDR isolates [48]. This variation could be attributable to fewer samples evaluated, samples were taken from intensive care unit patients, or that the study was conducted in a different geographic location and Biset et al [51]

Resistance patterns of Gram-negative bacilli to extended spectrum cephalosporins
The total resistance profile of Enterobacterales to extended-spectrum cephalosporins ranges from 57.7% to 80.8%, which marginally agrees with results of study conducted by Beyene et al 73.5% to 73.9% [37], and by Breurec S et al 79.7% [52] but higher than the study findings of Teklu et al (60.3% to 62.2%) [49], Gashe F et al (56.5% to 60.1%) [53] and Moges et al (60.3% to 66%, in 2021) [43] lower than the study finding by Moges et al (87.2% to 96.6%, in 2019) [44]. The geographical location, study time, and quantity of samples analyzed could all be factors in the reported discrepancy. Extended spectrum cephalosporin resistance was found in 17.5% of clinical Enterobacterales isolates studied in North America and Europe between 2016 and 2018 [54]; the observed discrepancy could be attributed to the study's large geographical scope, infection control techniques used in those settings, the number of samples tested, and other factors.

Conclusion and recommendations
This study revealed an alarming level of AMR, with a high prevalence of multidrug-resistant, extremely drug-resistant, carbapenemase-producing gram-negative bacteria, particularly among intensive care unit patients at the health facility level. These findings point to a scenario in which the clinical management of infected patients becomes increasingly difficult and necessitates the use of last-resort antimicrobials, likely exacerbating the magnitude of the global AMR crisis. This necessitates a robust antimicrobial resistance monitoring and infection prevention and control program at these institutions.

Limitations of the study
The responsible genes for carbapenemase production were not genotypically assessed. Further, patient clinical impact was not assessed. Tigecycline, colistin, and fosfomycin were not available and accordingly not used for AST.