Genetic Relatedness of NDM-Producing Klebsiella pneumoniae Co-Occurring VIM,KPC,and OXA-48 Enzymes from Surveillance Cultures from an Intensive Care Unit

Introduction

The presence of carbapenem-resistant Acinetobacter baumannii, carbapenem-resistant Pseudomonas aeruginosa, and mainly carbapenem-resistant Enterobacteriaceae (CRE) in health care facilities, and especially in the intensive care unit (ICU), constitutes a constant public health threat.^1–3 The intestinal tract represents an important reservoir for nosocomial pathogens, including CRE. Routine surveillance cultures for gastrointestinal screen to CRE have become a common practice and essential part of hospital infection prevention and control programs in many countries.^4–6 Therefore, intestinal carrier identification of multidrug-resistant (MDR) bacteria can be crucial to decide upon the adoption of contact precautions and other interventions to limit the spread of this pathogen.^7–10

Currently, MDR Klebsiella pneumoniae (MDR-KP) is considered one of the microorganisms most associated with nosocomial infections, especially in patients admitted to ICU. ¹¹ It has shown high rates of resistance to a broad spectrum of antimicrobials, including β-lactams, fluoroquinolones, and aminoglycosides.^12,13 Some studies have revealed the relationship of colonized patients with subsequent infection with MDR-KP.^14–17 Recently, Gorrie et al. revealed that K. pneumoniae is a fairly common component of the human gut microbiota (6%), and may have an impact on the health of ICU patients. Among 27 infections with this micro-organism diagnosed in the ICU, 48% occurred in patients considered previously colonized. ¹⁸ Furthermore, the MDR organism transmission between patients is an important mechanism for outbreaks that emerge in hospitals, especially in ICUs.^1,19,20 Thus, active surveillance is considered an important approach in the prevention and control of antimicrobial resistance programs, enabling early detection of emerging pathogens and epidemiological characteristic monitoring.^10,21

The K. pneumoniae carbapenemase (KPC; type A), Verona integron-encoded metallo-β-lactamase (VIM; type B), the New Delhi metallo-β-lactamase (NDM; type B), and the oxacillinase-48 (OXA; type D carbapenemase) are the most frequent carbapenemases with worldwide expansion. ²² The CRE harboring the bla_KPC-2 gene is among the most frequently isolated pathogen. This allele is usually carried by mobile genetic elements such as plasmids and transposons, which may be associated with other carbapenemase genes.^20,23–26 This way, it can spread horizontally between pathogenic and saprophytic bacteria, increasing the antibiotic resistance reservoir in both environmental and clinical microbiome. ²⁷ P. aeruginosa was the first micro-organism to carry VIM gene, but has quickly spread to Enterobacteriaceae.^28–31 VIM-encoding CRE is endemic in Greece and Italy, and their emergence in the United States may be associated with travel and hospitalization in these countries.^26,32,33 To date, 28 NDM enzyme variants have been found in Enterobacteriaceae, Acinetobacter spp., and P. aeruginosa from several countries^34,35 This carbapenemase has a greater ability to degrade penicillins, cephalosporins, and a few of the carbapenems compared to VIM, and a higher propagation potential than KPC.^36,37

Several reports have been suggesting the relationship of multidrug resistance with molecular epidemiological data based on multilocus sequence typing (MLST).^38–40 K. pneumoniae ST258 is the most common carbapenem-resistant ST type in the United States and Greece. It has been reported in >25 countries, including KPC epidemic countries such as Northeastern United States, Argentina, Brazil, Colombia, Eastern China, Greece, Israel, Italy, Poland, and Puerto Rico.^41,42 In addition, it is responsible for >77% of the outbreaks in the United States and 90% of all carbapenem-resistant K. pneumoniae (CRKP) infections in Israel.^43,44 Moreover, ST11 and ST437 are described as prevalent K. pneumoniae clones in Brazil. ⁴⁵

This study aimed to investigate the genetic relatedness of NDM-producing K. pneumoniae co-occurring with VIM, KPC, and OXA-48 enzymes in active surveillance cultures among patients admitted to ICU across the MLST scheme.

Materials and Methods

Results

Antimicrobial susceptibility, resistance genes, and MLST profiles

Out of the 631 colonies suggestive of CRKP, 108 isolates were phenotypically identified as carbapenemase-producing K. pneumoniae (99% identified by the Vitek II), 103 of these were confirmed as K. pneumoniae by 16S–23S rDNA internal transcribed spacer PCR. These 103 isolates showed resistance to all β-lactam antimicrobials (tazobactam/piperacillin, sulbactam/amoxicillin, cefepime, cefoxitin, ceftazidime, ceftriaxone, and cefuroxime), including carbapenems (ertapenem, imipenem, and meropenem), suggesting carbapenemase production. Ninety-four percent (97/103) of the isolates were resistant to ciprofloxacin, 84% (87/103) to tigecycline, 77% (79/103) to gentamicin, 62% (64/103) to amikacin, and 35% (36/103) to colistin.

Among the 103 isolates, 11% (11/103) showed the NDM gene and were grouped in three resistance profiles and four resistance genotypes (Table 1). Nine isolates of K. pneumoniae were classified as MDR, one as extensively drug resistant (XDR) and one as pandrug resistant (PDR) with high percentages of β-lactam resistance, including carbapenems, as well as ciprofloxacin 90% (10/11), tigecycline 82% (9/11), gentamicin 73% (8/11), and amikacin 18% (2/11) (Table 1). Furthermore, two isolates (P5335 and P5336) exhibited resistance to colistin, but did not carry mcr-related genes. It is worthy to note that one of these isolates (P5335) was resistant to all antimicrobials analyzed. Twenty-seven percent (3/11) of NDM-harboring K. pneumoniae revealed co-occurrence with KPC, OXA-48, and VIM type, 46% (5/11) with KPC and VIM, and 18% (2/11) with VIM. The bla_BKC, bla_SPM, and bla_IMP genes were not detected.

OPEN IN VIEWER

Table 1.

Characterization of NDM-Producing Klebsiella pneumoniae Strains

Strains	Age	Resistance		Gene profile		ST
		Phenotype	Profile
P5335	72 Years/male	AK; CIP; TIG; GEN; COL	PDR	NDM/VIM/KPC	III	4050 a
P5336	NI/female	CIP; TIG; GEN; COL	XDR	NDM/VIM/KPC	III	4051 a
P5337	64 Years/male	CIP; TIG	MDR	NDM/VIM/KPC/OXA-48	IV	515
P5339	7 Month/male	TIG; GEN	MDR	NDM/VIM/KPC	III	4054 a
P5430	2 Years/male	CIP; TIG; GEN	MDR	NDM	I	4047 a
P5435	2 Years/male	CIP; TIG; GEN	MDR	NDM/VIM	II	4048 a
P5462	84 Years/female	CIP; TIG; GEN	MDR	NDM/VIM/KPC	III	4049 a
P5678	37 Years/female	CIP; TIG	MDR	NDM/VIM/KPC	III	661
P5750	83 Years/female	CIP	MDR	NDM/VIM/KPC/OXA-48	IV	4052 a
P5763	76 Years/female	AK; CIP; GEN	MDR	NDM/VIM	II	15
P5773	73 Years/male	CIP; TIG; GEN	MDR	NDM/VIM/KPC/OXA-48	IV	4053 a

a

New ST.

AK, amikacin; CIP, ciprofloxacin; COL, colistin; GEN, gentamicin; MDR, multidrug resistant; NI, not informed; PDR, pandrug resistant; ST, sequence typing; TIG, tigecycline; XDR, extensively drug resistant.

MLST analysis of the 11 isolates harboring the bla_NDM gene revealed 11 different STs, 8 being (ST4047, ST4048, ST4049, ST4050, ST4051, ST4052, ST4053, and ST4054) novel and 3 (ST15, ST515, and ST661) previously described (Table 1). Two clonal complexes (CCs) were found in this study, including CC258 with two strains and CC15 with three strains. CC grouping was based on two different ST sharing at least 5 identical alleles (Table 2 and Fig. 1).

OPEN IN VIEWER

FIG. 1.

ST distribution among geographic location. CC, clonal complex. Color images are available online.

OPEN IN VIEWER

Table 2.

Clonal Complex Grouping Proposition of Strains Sharing Five or More Alleles Among ST

Discussion

Gastrointestinal colonization is considered a significant reservoir of CRKP, which has been considered a major public health problem. For this purpose, it is necessary to conduct active surveillance cultures in health centers, especially in ICU patients.^18,54 In addition, CRKP is an important bacterial pathogen responsible for hospital outbreaks worldwide and has been emerging in Brazilian hospitals. ⁵⁵

Our study showed that all 11 NDM-producing K. pneumoniae were resistant to other antimicrobial classes, including gentamicin, ciprofloxacin, and tigecycline. The high rates of aminoglycoside resistance exhibited by K. pneumoniae can be attributed to the excessive use of gentamicin, beyond the possibility of the presence of aac(3)-II gene in the same mobile genetic element carrying the carbapenemase genes. ⁵⁶ In contrast, another study reported a CRKP susceptible to different aminoglycosides, such as gentamicin, ⁵⁷ which could be related to the stability of plasmids in the presence or absence of selective pressure.^58,59 Regarding the treatment, the use of gentamicin alone or in combination with tigecycline has been used to reduce mortality from sepsis caused by CRE. ⁶⁰ Of note, in our study, the high percentages of CRKP were resistant to both gentamicin and tigecycline.

During the past decades, carbapenem resistance among K. pneumoniae has been typically caused by the emergence of transmissible carbapenemases, such as KPC and NDM. The NDM especially comprises one of the most rapidly growing group of metallo-β-lactamases.^36,61 They have been increasingly detected in different countries, suggesting a worldwide dissemination.^56,61 Thus, carbapenem antimicrobials are becoming increasingly ineffective and only a few therapeutic options are available, such as colistin. ⁶²

According to Nordmann et al., the majority of NDM-producing bacteria also carry a diversity of other resistance mechanisms, including ESBLs (especially CTX-M-15) and different carbapenemases (e.g., OXA-48, VIM, and KPC types). ⁶³ The coexistence phenomenon of multidrug resistance determinants in a single strain has become commonplace.^64–66 On the other hand, another study that evaluated the genetic relationship and molecular determinants of carbapenem resistance in clinical strains, revealed that none of the 100 CRKP harbored either KPC or VIM genes. Furthermore, they exhibited a prevalence of OXA-48 and NDM, as well as an increase in the colistin resistance rate. ⁶⁷ The existence of at least two carbapenemase genes in CRKP also has been reported worldwide, including KPC and VIM in Italy,^1,26,68 KPC-2 and VIM-24 in Colombia, ⁶⁹ NDM-1 and KPC-2 in Brazil, ⁷⁰ and NDM-1 and OXA-181 in Singapore. ⁷¹ Other studies generally characterize the CRKP carrying a single carbapenemase. However, we report for the first time K. pneumoniae co-harboring up to four carbapenemase genes (NDM, KPC, OXA-48, and VIM) from surveillance cultures belonging to ST515.

Molecular epidemiology data based on MLST showed the variety of K. pneumoniae ST associated with drug resistance genes, such as ST258, which is the most common carbapenem-resistant genotype in the United States, Brazil, Italy, Norway, and Greece^55,72,73 (Fig. 1). A study carried out in Russia also showed that ST340 was associated with K. pneumoniae producing NDM. ⁷⁴ Meanwhile, recent evidence showed that global epidemiology is changing, with the rise of new lineages such as KpV_S_1 and KpV_S_2 isolates belonging to ST661 and ST307, respectively. Both STs have been recently reported as responsible for outbreaks in several European countries. Furthermore, the detection of bla_VIM-1 gene in nonpredominant K. pneumoniae clones in a hospital setting is strong evidence of the changing epidemiology in Italy. ⁷⁵ In Argentina, ST258 is no longer the absolute clone among KPC-Kp isolates with the rise of more virulent lineages such as the hypermucoviscous ST25 and the emergence of the high-risk clone ST307. ⁷⁶

In this study, the 11 NDM-producing K. pneumoniae were grouped into 11 different STs, with 8 new and 3 previously described (ST515, ST661, and ST15). Six of the eight new STs were associated with MDR strains, in addition to a PDR isolate belonging to ST4050. According to the MLST database, ST661 is a rare lineage isolated only in China and United States, and has been implicated in an outbreak in the United Kingdom, which involved both the clonal spread of a KPC-producing K. pneumoniae and the dissemination of pKpQIL-D2-KPC-2 plasmid. ⁷⁷ In turn, ST15 is a worldwide spread genotype, but has never been reported as a source of hospital outbreaks of CRKP. As previously reported,^78–81 the ST15 strain isolated in this study presented an MDR profile associated with VIM and NDM genes, suggesting a high capacity of horizontal transfer of resistance genes. ⁸² Indeed, whole-genome sequencing of K. pneumoniae ST15 strains revealed high antimicrobial resistance sharing a common ancestor with 140 globally disseminated K. pneumoniae strains. ⁸³ Another genotype found in this study, ST515, has been reported in Poland and Vietnam as a sporadic infection agent. It is worthy of note that ST515 and ST15 share 5 MLST alleles (Table 2), suggesting that these two STs may be grouped into the same CC15 that is settling among healthy carriers and patients either in the hospital environment or outside health facilities. The STs 4051 and 4054 are closely related to antimicrobial resistance-associated ST258, sharing 5 alleles, and could be grouped into CC258 (Table 2). Strains from these two STs harbored bla_KPC, bla_NDM, and bla_VIM genes, besides high resistance to β-lactams, ciprofloxacin, tigecycline, and gentamicin (Table 1).

Conclusion

In conclusion, our data reveal new multiresistant CRKP clones from surveillance cultures that may represent a concern for health authorities regarding the spread of antibiotic resistance. Notably, of the eight new STs described in this study, two are resistant to colistin with no mcr allele (1 to 8), but showing co-production of KPC, NDM, and VIM. Indeed, the presence of NDM in the 11 different STs may reflect the possibility of plasmid transfer carrying several carbapenemases genes, as has been reported worldwide. This event can have serious consequences for public health, leading to an increased spread of antimicrobial genes and MDR microorganisms turning treatment against bacterial pathogens difficult.

The active surveillance culture program could be considered an essential approach to fight the threat of antimicrobial resistance. Moreover, the disclosure of our data must be shared both with the scientific community and with the main stakeholders, including the population and government decision makers.