Introduction

Antimicrobial resistance (AMR) among Gram-negative pathogens (GNP) increased worldwide. A high rate of AMR has been reported in Egypt since more than 20 years, among GNP causing nosocomial infections and outbreaks [1,2,3]. AMR rates have increased especially among nosocomial GNP, probably due to widespread abuse of antimicrobials including carbapenems in Egyptian hospitals and poor compliance with infection control practices [4, 5]. In small-scale studies, blaOXA, blaNDM, blaVIM, blaIMP, and blaKPC carbapenemase genes were detected in Egypt [6,7,8]. As the genetic basis of beta-lactam resistance was not yet studied at a large scale in Egypt, we aimed to molecularly characterize multidrug-resistant (MDR) GNP.

Methods

Study sites and strains

This study was conducted in 3 major tertiary care Egyptian hospitals participating in the “Study for Monitoring Antimicrobial Resistance Trends” (SMART) from 2015 to 2016. The hospitals were Ain Shams University Hospital, Ain Shams Specialized University Hospital, and Dar Al-Fouad Hospital. Isolates were collected according to SMART protocol as previously reported [9,10,11]. Briefly, the participating hospitals collected 1070 non-repetitive consecutive GNP isolates from lower respiratory tract specimens, urine and pus or abdominal fluid of hospitalized patients with lower respiratory tract infections (LRTI), complicated urinary tract infections (cUTI), and complicated intra-abdominal infections (cIAI) during the study period. Of these, we studied 495 isolates that showed phenotypic resistance to third-generation cephalosporins or carbapenems. The identification of GNP, susceptibility testing, and detection of resistance phenotypes were conducted in hospital laboratories according to the methods of the Clinical and Laboratory Standards Institute [12] and confirmed in a reference laboratory (International Health Management Associates (IHMA), Inc., Schaumburg, IL, USA), where susceptibility and extended-spectrum β-lactamase (ESBL) phenotype were determined using the CLSI broth microdilution method with custom dehydrated panels manufactured by Trek Diagnostic Systems (Thermo Scientific, Independence, OH) in 2015–2016. MIC interpretive criteria followed the 2017 M100-S27 guidelines of the CLSI [13]. EUCAST breakpoints were used only for colistin against Enterobacteriaceae, because no CLSI breakpoints exist [14].The susceptibility of all Gram-negative isolates combined was calculated using breakpoints appropriate for each species.

Genotypic identification of antimicrobial resistance genes

The molecular characterization of ESBL and carbapenemases was done using the Check-Points MDR CT103 (Check-Points Health B.V., Wageningen, The Netherlands) microarray kit, which detects most carbapenemase, ESBL, and AmpC genes: ESΒLs (class A)—blaTEM, blaSHV, blaCTX-M, blaVEB, blaPER, and blaGES; blaAmpC β-lactamase genes (class C)—blaACC, blaACT, blaCMY, blaDHA, blaFOX, blaMIR, and blaMOX; and carbapenemases (class A)—blaKPC and blaGES; (class B)—blaNDM, blaIMP, blaVIM, blaGIM, and blaSPM; and (class D)—blaOXA-48-like. Then the genes encoding ESΒL, carbapenemases, and AmpC were sequenced in their entirety in IHMA [11].

Results

Strains and phenotypic antibiotic susceptibility

The 495 MDR isolates were derived from cIAI (181, 36%), LRTI (128, 26%), and cUTI (186, 38%). Overall Klebsiella pneumoniae and Escherichia coli were the most common (52.7% and 44.2%) and were also the predominant organisms in cIAI (52.2% and 45.9%, respectively), cUTI (43% and 54.8%, respectively), and LRTI (71.1% and 26.6%, respectively) (Table 1). K. pneumoniae was most susceptible to colistin, amikacin, ertapenem, and imipenem (92.7%, 72.7%, 69.3%, and 64%, respectively). E. coli remained most susceptible to colistin (100%), amikacin (94.1%), imipenem (90.4%), and ertapenem (83.6%).

Table 1 Distribution of microorganism among clinical isolates
Full size table

Identification of blaESBL genes of the TEM, SHV, and CTX-M types

ESBL production was detected in 96.2% of the MDR isolates (Table 2). ESBL genotypes included blaCTX-M-15 (70.1%), blaTEM-OSBL (48.5%), blaSHV-OSBL (27.9%), and blaCTX-M-14 (10.7%). The predominant ESBL gene in both E. coli and K. pneumoniae was blaCTX-M-15 (Table 3).

Table 2 Percentage of ESΒL, AmpC β-lactamases, and carbapenemase genes among 495 MDR isolates
Full size table
Table 3 ESΒL predominant genotypes among E. coli and K. pneumoniae
Full size table

Identification of carbapenemase genes

Carbapenem resistance genes were detected in 45.3% of the MDR isolates. In K. pneumoniae, blaOXA-48 dominated (40.6%), followed by blaNDM-1 (23.7%) and blaOXA-232 (4.5%). In E. coli, the most frequent genes were blaNDM-5 (9.6%), blaOXA-181 (5.5%), blaOXA-244 (3.7%), and blaNDM-1 (3.7%). blaKPC-2 and blaVIM-2 were less frequently identified (Table 4).

Table 4 Carbapenemase genotypes among E. coli and K. pneumoniae
Full size table

Identification of AmpC β-lactamases resistance genes

AmpC resistance genes were identified in 9.7% of the isolates; blaCMY-2 was the most predominant one (Table 5). In 153 isolates (32.3%), coexistence of 2 or more resistance genes was detected (Table 4). The commonest combination of 2 genes was blaCTX-M-15 with blaNDM-5 (2.6%); the commonest combination of 3 genes was blaSHV-OSBL, blaCTX-M-15, and blaNDM-1 (9.2%); the commonest combination of 4 genes was blaSHV-OSBL, blaTEM-OSBL, blaCTX-M-15, and blaNDM-1 (12.4%); and the commonest combination of more than 4 genes was blaSHV-OSBL, blaTEM-OSBL, blaCTX-M-15, blaCTX-M-14, blaNDM-1, and blaOXA-48 (5.9%) (Table 6). Tables 7 and 8 show K. pneumoniae and E.coli susceptibility (%) against MDR- GNP per type of infection.

Table 5 Total AmpC β-lactamases genotype among Enterobacteriaceae isolates
Full size table
Table 6 Common combinations of resistant genes among tested isolates
Full size table
Table 7 K. pneumoniae susceptibility (%) against MDR-GNP per type of infection
Full size table
Table 8 E. coli susceptibility (%) against MDR-GNP per type of infection
Full size table

Discussion

Surveillance for AMR is essential to monitor trends, identify emerging resistance mechanisms, and support the antimicrobial stewardship programs. To our knowledge, this is the most extensive molecular study of AMR in Egypt including 495 MDR Gram-negative isolates. In all hospitals, more than half of infections were present on admission (community acquired or transferred from other hospitals; data not shown). We identified 43 resistance phenotypes distributed among isolates from the 3 hospitals. Typing of nosocomial GNP in each hospital based on the phenotypic resistance patterns showed no clonal spread except for few isolates in each hospital (data not shown).

K. pneumoniae was a common pathogen in all 3 types of infection, and most isolates were MDR. Due to its large accessory genome including plasmids and chromosomal loci, K. pneumoniae isolates may act as opportunistic pathogens. Such strains infect critically ill and immunocompromised patients mostly, whereas other strains of K. pneumoniae (hyper-virulent) may even infect healthy people in community settings. Many of the virulent strains encode carbapenemases [15]. E. coli was a frequently identified pathogen from cIAI and cUTI, which is consistent with previous reports, and was also a frequent pathogen in LRTI. E. coli pneumonia is uncommon and may result from micro-aspiration of colonized upper airway secretions in severely ill patients; hence, it is a well-known cause of nosocomial pneumonia [16]. However, E. coli pneumonia may also be community-acquired in patients who have underlying diseases such as diabetes mellitus, alcoholism, chronic obstructive pulmonary disease, and E. coli UTI [17]. Our results confirm the results of previous studies in Egypt [3, 18] that showed high rates of ESΒL and carbapenem resistance among ICU pathogens. For example, in the study of Talaat et al. [18], ESBL and carbapenem resistance were identified in 54% and 13.8% of E. coli isolates compared with 42.5% and 48.1% in K. pneumoniae isolates, respectively.

Among the many types of ESΒLs reported, blaCTX-M-15 and blaCTX-M-14 are the most commonly identified worldwide as the genes encoding CTX-M enzymes (blaCTX-M) can be horizontally mobilized by various genetic elements [19]. This reflects the situation in Egypt as well. AmpC genes were less frequently identified (9.7%), usually in combination with other resistance genes. CMY-2-like enzymes were the most predominant. Other detected genes were blaCMY-42, blaDHA, and blaACT-like. This is in keeping with recent reports on acquisition of plasmid-mediated cephalosporinase producing Enterobacteriaceae after a travel to the tropics and North Africa including Egypt [20, 21]. The high rate of ESΒL in Egypt is essentially due to inappropriate use of antimicrobials in human and animal health care. Of the patients treated in public outpatient health care facilities, 49.8% received antibiotics, and antibiotics are still sold over the counter without prescription [22]. Moreover, there is extensive use of antimicrobials for prevention and treatment of infections in veterinary care, and ESΒL and AmpC resistance mechanisms were detected in veterinary E. coli isolates in Egypt [23]. This is an urgent public health problem especially with a growing body of evidence supporting foodborne transmission of resistance, especially from poultry, as poultry meat exhibits the highest levels of contamination by MDR bacteria [24, 25].

The most important mechanism of carbapenem resistance is the production of carbapenemases; therefore, all isolates were investigated to identify the carbapenemase genes. We identified them in 45.4% of tested isolates; blaOXA followed by blaNDM genes dominated, while only 2 isolates of K. pneumoniae harbored blaKPC-2 genes. These results confirm previous small-scale reports that the blaNDM and blaOXA genes are the predominant in Egypt and Middle East [26].

The present study detected massive coexistence of different resistance genes among tested isolates. This coexistence could have contributed to the observed elevated variability in resistance phenotypes and genotypes among GNP in Egypt [15].

In conclusion, our study detected alarming rates of resistance and identified many resistance mechanisms in clinical GNP from Egyptian tertiary care hospitals. These high resistance rates highlight the importance of continuous monitoring of the resistance trends, adherence to infection control policies, and underscore urgently implementing a national antimicrobial stewardship plan in Egypt.