Global Dissemination of bla_KPC into Bacterial Species beyond Klebsiella pneumoniae and In Vitro Susceptibility to Ceftazidime-Avibactam and Aztreonam-Avibactam

Abstract

The Klebsiella pneumoniae carbapenemase (KPC), first described in the United States in 1996, is now a widespread global problem in several Gram-negative species. A worldwide surveillance study collected Gram-negative pathogens from 202 global sites in 40 countries during 2012 to 2014 and determined susceptibility to β-lactams and other class agents by broth microdilution testing. Molecular mechanisms of β-lactam resistance among carbapenem-nonsusceptible Enterobacteriaceae and Pseudomonas aeruginosa were determined using PCR and sequencing. Genes encoding KPC enzymes were found in 586 isolates from 22 countries (76 medical centers), including countries in the Asia-Pacific region (32 isolates), Europe (264 isolates), Latin America (210 isolates), and the Middle East (19 isolates, Israel only) and the United States (61 isolates). The majority of isolates were K. pneumoniae (83.4%); however, KPC was detected in 13 additional species. KPC-2 (69.6%) was more common than KPC-3 (29.5%), with regional variation observed. A novel KPC variant, KPC-18 (KPC-3[V8I]), was identified during the study. Few antimicrobial agents tested remained effective in vitro against KPC-producing isolates, with ceftazidime-avibactam (MIC₉₀, 4 μg/ml), aztreonam-avibactam (MIC₉₀, 0.5 μg/ml), and tigecycline (MIC₉₀, 2 μg/ml) retaining the greatest activity against Enterobacteriaceae cocarrying KPC and other β-lactamases, whereas colistin (MIC₉₀, 2 μg/ml) demonstrated the greatest in vitro activity against KPC-positive P. aeruginosa. This analysis of surveillance data demonstrated that KPC is widely disseminated. KPC was found in multiple species of Enterobacteriaceae and P. aeruginosa and has now become a global problem.

INTRODUCTION

Infections caused by carbapenem-resistant Enterobacteriaceae (CRE) contribute to attributable mortality higher than that for patients infected with carbapenem-susceptible isolates (1). The effect of CRE on morbidity and mortality can vary significantly between countries and may depend upon the β-lactam resistance mechanisms that are most problematic in certain regions (2–5). Population movements, poor infection control, and the lack of antimicrobial stewardship initiatives have perpetuated the dissemination of genes that encode carbapenemases among clinically significant bacterial species on a global scale (2, 4, 6, 7). Detection of CRE and their associated resistance mechanisms is essential in order to determine the appropriate therapeutic options required for a positive patient infection outcome (8–10).

The Klebsiella pneumoniae carbapenemase (KPC) is a class A serine carbapenemase first recognized in the northeastern United States in 1996 (11). Bacterial pathogens expressing KPC are clinically significant in that they are often multi- or pan-drug resistant, including resistance to currently available latest-in-line therapeutic options (7, 12, 13). The impact of KPC became more fully recognized as this family of enzymes became a global threat to public health, in that the gene encoding KPC (bla_KPC) has now been observed in multiple Enterobacteriaceae species and has disseminated worldwide, in large part due to the spread of K. pneumoniae isolates belonging to the successful high-risk clonal complex 258 (7, 13). bla_KPC is most often embedded within the Tn4401 transposon, though it has also been reported in other mobile elements, and found in plasmids belonging to 12 incompatibility groups capable of species-to-species transfer within Enterobacteriaceae and some nonfermentative Gram-negative pathogens, including Pseudomonas aeruginosa (14–17). Furthermore, these plasmids commonly also carry genes encoding aminoglycoside resistance mechanisms and additional β-lactamases, including extended-spectrum β-lactamases (ESBLs) (12, 17). bla_KPC has also been found inserted into the bacterial chromosome (18, 19).

This investigation documented the distribution of KPC-producing Gram-negative bacterial pathogens isolated from a sampling of clinically significant pathogens collected during a global surveillance study.

MATERIALS AND METHODS

Nonduplicate, nonconsecutive isolates from intra-abdominal, urinary tract, skin and soft tissue, lower respiratory tract, and bloodstream infections were collected from 202 medical centers in 40 countries located in the Asia-Pacific region, Europe, Latin America, the Middle East-Africa, and North America. The medical centers were instructed to contribute a specific number of isolates of each requested species regardless of antibiotic susceptibility. Participating countries by year are listed in Table S1 in the supplemental material. Basic patient demographic data were collected but were not linked to patient identity or therapeutic outcome. Organism collection, transport, confirmation of organism identification, susceptibility testing, molecular characterization, data quality assurance, and development and management of a centralized database were coordinated by a central laboratory.

Matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry (Bruker Daltronics, Bremen, Germany) was used to confirm the organism identification of all isolates. Antibiotic susceptibility testing was performed by broth microdilution using custom frozen panels. Ceftazidime-avibactam and aztreonam-avibactam were tested at a fixed concentration of 4 μg/ml avibactam. Panel manufacture, inoculation, incubation, interpretation, and quality control testing were performed according to Clinical and Laboratory Standards Institute (CLSI) guidelines (20, 21). All P. aeruginosa isolates that were nonsusceptible to doripenem, meropenem, or imipenem and Enterobacteriaceae isolates that were nonsusceptible to those carbapenems or ertapenem using CLSI breakpoints were molecularly characterized for β-lactamase genes encoding KPC and other β-lactamases (OXA-48-like, TEM, SHV, CTX-M, VEB, PER, GES, ACT, CMY, DHA, MIR, ACC, MOX, FOX, NDM, IMP, VIM, SPM, and GIM) using a combination of microarray and multiplex PCR assays, followed by sequencing as previously described (22).

Nucleotide sequence accession number.

The sequence of the new variant KPC-18 was deposited in GenBank with accession no. KP681699.

RESULTS

A total of 38,266 isolates of Enterobacteriaceae and 8,010 isolates of P. aeruginosa were collected in 40 countries participating in a global surveillance study in 2012 to 2014. Of these, 586 (1.3%) carbapenem-nonsusceptible Gram-negative isolates collected from medical centers in 22 countries carried bla_KPC. In addition, four carbapenem-susceptible KPC-positive isolates were also identified as part of the study but were excluded from analysis.

With the exception of 34 isolates collected during patient visits to an emergency room, all carbapenem-nonsusceptible, KPC-positive isolates were from patients admitted to an inpatient hospital ward, with one-third collected in intensive care units (data not shown). Patient ages ranged from <1 to 95 years, with a median age of 61 years, and more male patients (354, 60.4%) were identified with a KPC-positive isolate than females (220, 37.5%); information regarding gender was not available for 12 patients. KPC-positive bacteria were isolated from various infection sources, with the greatest number of overall isolates (189 of 586, 32.2%) collected from respiratory cultures and approximately equal numbers of isolates collected from urinary tract infections (137) and skin and soft tissue infections (132). The largest numbers of isolates were collected from urine (126), wounds (71), endotracheal aspirates (67), sputum (63), and peritoneal fluid (45) (Table 1). In cases when information regarding the length of hospital stay was available, 75.1% of patients with a KPC-positive isolate were admitted more than 48 h prior to culture, suggesting nosocomial acquisition (Table 1).

TABLE 1.

Body source distribution of 586 carbapenem-nonsusceptible KPC-positive Enterobacteriaceae and P. aeruginosa isolates collected in 2012 to 2014

Source of infectiona	Source of culture	No. of isolates from patients hospitalized for:
		<48 h	≥48 h	NAb
Gastrointestinal tract	Peritoneal fluid	12	33
	Abscess	3	25
	Gallbladder	3	11
	Otherc	2	12
	Total (%)	20 (3.4)	81 (13.8)
Urinary tract	Urine	41	85
	Ureter/urethra/bladder	3	5
	Otherd		3
	Total (%)	44 (7.5)	93 (15.9)
Skin and soft tissue	Wound	22	48	1
	Decubitus ulcer	7	22
	Abscess	2	12
	Carbuncle/furuncle/cellulitis/erysipelas	2	6
	Burn	3	4
	Othere		3
	Total (%)	36 (6.1)	95 (16.2)	1 (0.2)
Respiratory tract	Endotracheal aspirate	14	51	2
	Sputum	14	48	1
	Bronchoalveolar lavage fluid	5	27	1
	Bronchial brushing	3	10
	Otherf	4	9
	Total (%)	40 (6.8)	145 (24.7)	4 (0.7)
Bloodstream	Blood	1	25
	Total (%)	1 (0.2)	25 (4.3)
Unknown	Unknown		1
	Total (%)		1 (0.2)
Total (%)		141 (24.1)	440 (75.1)	5 (0.9)

^aSpecies found (number of isolates): gastrointestinal tract, C. farmeri (1), C. freundii (3), C. koseri (1), E. aerogenes (1), E. cloacae (3), E. coli (4), K. oxytoca (3), K. pneumoniae (84), and P. aeruginosa (1); urinary tract, C. freundii (2), E. cloacae (3), E. coli (10), K. oxytoca (3), K. pneumoniae (110), and P. aeruginosa (9); skin and soft tissue, C. amalonaticus (1), C. freundii (1), E. asburiae (2), E. cloacae (2), E. coli (8), K. oxytoca (3), K. pneumoniae (111), and P. aeruginosa (4); respiratory tract, C. freundii (1), C. koseri (1), E. aerogenes (2), E. asburiae (1), E. coli (2), K. oxytoca (4), K. pneumoniae (159), M. morganii (1), P. aeruginosa (15), R. ornithinolytica (1), and S. marcescens (2); bloodstream, E. cloacae (1), K. oxytoca (1), and K. pneumoniae (24); unknown, K. pneumoniae (1).

^bNA, not available because the hospital admission date was not provided by the investigator.

^cOther sources (number of isolates): appendix (4), liver (4), large colon (3), not specified (2), and pancreas (1).

^dOther sources (number of isolates): not specified (2), and prostate (1).

^eOther sources (number of isolates): not specified (3).

^fOther sources (number of isolates): thoracentesis (5), not specified (5), and lungs (3).

K. pneumoniae was the most commonly isolated KPC-producing species (n = 489, 83.4%), followed by P. aeruginosa (29, 4.9%), Escherichia coli (24, 4.1%), and Klebsiella oxytoca (14, 2.4%). The remaining 5% of KPC-positive isolates were composed of 10 species of Enterobacteriaceae (9 Enterobacter cloacae, 7 Citrobacter freundii, 3 each Enterobacter aerogenes and Enterobacter asburiae, 2 each Citrobacter koseri and Serratia marcescens, and 1 each Citrobacter amalonaticus, Citrobacter farmeri, Morganella morganii, and Raoultella ornithinolytica) (Table 2). KPC-positive isolates were collected in all regions, with large numbers of isolates collected in countries in which KPC-producing organisms were previously reported to be endemic (Greece, Italy, Colombia, Argentina, Brazil, the United States, China, and Israel), as well as Portugal, in which KPC-positive isolates have been reported only recently (Table 2; Fig. 1) (7, 23). It should be noted that isolates from patients in China were obtained only during 2012 to 2013 due to export restrictions of bacterial pathogens imposed in 2014.

TABLE 2.

Geographic and species distribution of 586 carbapenem-nonsusceptible KPC-positive isolates collected as part of a global surveillance program (2012 to 2014)

Region	Country	n	Organism	KPC variant(s) (n)
Europe	Austria	5	E. aerogenes	KPC-2 (1)
			K. oxytoca	KPC-2 (1)
			K. pneumoniae	KPC-2 (2), KPC-3 (1)
	Belgium	4	K. pneumoniae	KPC-2 (1), KPC-3 (3)
	Czech Republic	1	K. pneumoniae	KPC-2 (1)
	Germany	1	K. pneumoniae	KPC-2 (1)
	Greece	134	C. amalonaticus	KPC-2 (1)
			K. oxytoca	KPC-2 (3)
			K. pneumoniae	KPC-2 (128), KPC-9 (2)
	Italy	87	E. coli	KPC-2 (1)
			K. oxytoca	KPC-3 (1)
			K. pneumoniae	KPC-2 (15), KPC-3 (70)
	Portugal	24	C. freundii	KPC-3 (1)
			E. coli	KPC-3 (3)
			K. oxytoca	KPC-3 (1)
			K. pneumoniae	KPC-3 (19)
	Romania	6	K. pneumoniae	KPC-2 (6)
	Russia	1	E. cloacae	KPC-2 (1)
	United Kingdom	1	K. pneumoniae	KPC-2 (1)
Latin America	Argentina	60	E. cloacae	KPC-2 (2)
			E. coli	KPC-2 (2)
			K. oxytoca	KPC-2 (2)
			K. pneumoniae	KPC-2 (50), KPC-3 (2)
			M. morganii	KPC-2 (1)
			P. aeruginosa	KPC-2 (1)
	Brazil	60	E. cloacae	KPC-2 (2)
			E. coli	KPC-2 (1)
			K. oxytoca	KPC-2 (2)
			K. pneumoniae	KPC-2 (53), KPC-3 (2)
	Chile	17	K. pneumoniae	KPC-2 (1)
			P. aeruginosa	KPC-2 (16)
	Colombia	60	C. freundii	KPC-2 (3)
			C. koseri	KPC-2 (1)
			E. cloacae	KPC-2 (2)
			E. coli	KPC-2 (4)
			K. oxytoca	KPC-2 (1)
			K. pneumoniae	KPC-2 (24), KPC-3 (13)
			S. marcescens	KPC-2 (1)
			P. aeruginosa	KPC-2 (11)
	Mexico	2	K. pneumoniae	KPC-3 (1)
			R. ornithinolytica	KPC-3 (1)
	Venezuela	11	C. freundii	KPC-2 (1)
			K. pneumoniae	KPC-2 (10)
North America	United States	61	C. farmeri	KPC-3 (1)
			C. freundii	KPC-2 (1)
			E. asburiae	KPC-2 (2)
			E. cloacae	KPC-2 (1)
			E. coli	KPC-3 (3), KPC-18 (2)
			K. pneumoniae	KPC-2 (11), KPC-3 (40)
Asia-Pacific	Chinaa	28	C. koseri	KPC-2 (1)
			E. aerogenes	KPC-2 (1)
			E. cloacae	KPC-2 (1)
			E. coli	KPC-2 (5)
			K. oxytoca	KPC-2 (3)
			K. pneumoniae	KPC-2 (14), KPC-12 (1)
			S. marcescens	KPC-2 (1)
			P. aeruginosa	KPC-2 (1)
	Japan	1	E. asburiae	KPC-2 (1)
	Philippines	2	K. pneumoniae	KPC-2 (2)
	Taiwan	1	K. pneumoniae	KPC-2 (1)
Middle East-Africa	Israel	19	C. freundii	KPC-2 (1)
			E. aerogenes	KPC-2 (1)
			E. coli	KPC-2 (2), KPC-3 (1)
			K. pneumoniae	KPC-2 (4), KPC-3 (10)

^aNo isolates were obtained from patients in mainland China in 2014 due to export restrictions.

FIG 1.

Distribution of KPC-positive Enterobacteriaceae and P. aeruginosa collected in 2012 to 2014.

Five KPC sequence variants were identified, with 99.1% of isolates carrying either KPC-2 (408, 69.6%) or KPC-3 (173, 29.5%). KPC-2 was detected in 20 of 22 countries in this investigation, whereas KPC-3 was detected in 10 countries and was the only variant found in isolates collected in Mexico and Portugal. Larger proportions of isolates from Italy (81.6%), Israel (57.9%), and the United States (72.1%) carried KPC-3 in comparison to KPC-2. A total of 93.1% of detected KPC-3 variants were carried by K. pneumoniae. In contrast, KPC-2 was found in more diverse Gram-negative species, but still 79.7% were carried by K. pneumoniae. All KPC-positive P. aeruginosa isolates carried the KPC-2 variant, and all but one were collected from countries in Latin America (Table 2). Two K. pneumoniae isolates collected in Greece carried KPC-9, and one K. pneumoniae isolate collected in China carried KPC-12. One novel variant, KPC-18 (KPC-3[V8I]), was identified during this study. KPC-18 was detected in E. coli collected from two different patients within a 2-week period in 2014 at a medical center located in suburban Chicago, IL, USA. The first isolate was cultured from a patient with a respiratory infection in an intensive care unit, whereas the second isolate was collected from peritoneal fluid during an emergency room visit. Both isolates were resistant to ampicillin, aztreonam, ceftazidime, cefepime, doripenem, imipenem, meropenem, piperacillin-tazobactam, and levofloxacin but were susceptible in vitro to ceftazidime-avibactam, amikacin, tigecycline, and colistin and showed low MIC values (0.12 μg/ml) of aztreonam-avibactam (data not shown).

A total of 96.8% of KPC-positive isolates also carried additional β-lactamases, including plasmid-encoded and presumed intrinsic chromosomally encoded enzymes. Notably, nine isolates carried a second carbapenemase belonging to Ambler class B or class D. Four K. pneumoniae isolates collected in Greece carried KPC-2 and VIM-1, two K. pneumoniae isolates collected in China carried KPC-2 and IMP-4, one P. aeruginosa isolate collected in Chile carried KPC-2 and VIM-2, and two K. pneumoniae isolates collected in Greece and Argentina carried KPC-2 and OXA-163. Of these, 7 isolates also carried ESBLs and/or AmpC β-lactamases (Table 3). The majority (291, 49.7%) of isolates carried KPC alone or with an original-spectrum β-lactamase (OSBL) (TEM-1, TEM-2, SHV-1, or SHV-11) that is not expected to significantly impact susceptibility to antimicrobial agents in clinical use, with 36% of KPC-2-positive (n = 147) and 82.1% of KPC-3-positive (n = 142) isolates found in this subset. Approximately one-third of isolates (219, 37.4%) coproduced KPC and ESBLs, whereas isolates carrying KPC and AmpC β-lactamases (25, 4.3%) or KPC plus both AmpC and one or more ESBLs (14, 2.4%) were less frequently encountered. KPC-3 was most often found with CTX-M-15 (11 isolates) or SHV-12 (15 isolates) and was not found in combination with a plasmid-mediated AmpC in any of the isolates. KPC-2 was most often cocarried with SHV-12 (104 isolates), CTX-M-15 (35 isolates), and VEB-1 (18 isolates); the last combination was detected in 17 K. pneumoniae isolates and one E. aerogenes isolate collected from Greece (16 isolates) and Austria (2 isolates) (Table 3).

TABLE 3.

Cocarriage of KPC and other β-lactamases in carbapenem-nonsusceptible Enterobacteriaceae and P. aeruginosa collected in 2012 to 2014

β-Lactamasesa	Organism	n	Molecular variant(s)
KPC + MBL	K. pneumoniae	1	KPC-2, VIM-1
KPC + MBL + ESBL + AmpC + OSBL	K. pneumoniae	2	KPC-2, VIM-1, SHV-12, CMY-13, TEM-OSBL
KPC + MBL + ESBL	K. oxytocab	2	KPC-2, IMP-4, SHV-12
KPC + MBL + AmpC ± OSBL	K. pneumoniae	1	KPC-2, VIM-1, MOX-1, SHV-OSBL
	P. aeruginosac	1	KPC-2, VIM-2
KPC + ESBL-like OXA + ESBL + OSBL	K. pneumoniae	1	KPC-2, OXA-163, CTX-M-2, SHV-OSBL, TEM-OSBL
KPC + ESBL-like OXA + OSBL	K. pneumoniae	1	KPC-2, OXA-163, SHV-OSBL, TEM-OSBL
KPC + ESBL + AmpC ± OSBL	C. freundiic	1	KPC-2, SHV-12, TEM-OSBL
		1	KPC-2, CTX-M-15
		1	KPC-3, CTX-M-9, SHV-12
	C. koseric	1	KPC-2, CTX-M-3, TEM-OSBL
	E. aerogenesc	1	KPC-2, VEB-1, TEM-OSBL
	E. asburiaec	2	KPC-2, SHV-30, TEM-OSBL
	E. cloacaec	1	KPC-2, SHV-12, DHA-1
		3	KPC-2, CTX-M-15, TEM-OSBL
	K. pneumoniae	2	KPC-2, CTX-M-15, MOX-2, SHV-OSBL, TEM-OSBL
		1	KPC-2, CTX-M-27, DHA-1, SHV-OSBL
KPC + ESBL ± OSBL	E. coli	1	KPC-2, CTX-M-15
		1	KPC-2, CTX-M-15, TEM-OSBL
	K. oxytocab	4	KPC-2
		2	KPC-2, TEM-OSBL
		1	KPC-2, SHV-5, TEM-OSBL
		1	KPC-2, SHV-12
		1	KPC-2, CTX-M-8, TEM-OSBL
		1	KPC-2, CTX-M-15, TEM-OSBL
		2	KPC-3, TEM-OSBL
	K. pneumoniae	1	KPC-2, SHV-5, TEM-OSBL
		29	KPC-2, SHV-12
		64	KPC-2, SHV-12, TEM-OSBL
		1	KPC-2, SHV-12, CTX-M-14, TEM-OSBL
		2	KPC-2, SHV-12, CTX-M-65
		1	KPC-2, SHV-12, CTX-M-65, TEM-OSBL
		1	KPC-2, SHV-28, CTX-M-15, TEM-OSBL
		3	KPC-2, CTX-M-2, SHV-OSBL
		1	KPC-2, CTX-M-2, TEM-OSBL
		9	KPC-2, CTX-M-2, SHV-OSBL, TEM-OSBL
		1	KPC-2, CTX-M-2, CTX-M-15, SHV-OSBL, TEM-OSBL
		2	KPC-2, CTX-M-3, SHV-OSBL, TEM-OSBL
		1	KPC-2, CTX-M-12
		2	KPC-2, CTX-M-12, SHV-OSBL
		1	KPC-2, CTX-M-14, TEM-OSBL
		9	KPC-2, CTX-M-14, SHV-OSBL, TEM-OSBL
		10	KPC-2, CTX-M-15, SHV-OSBL
		14	KPC-2, CTX-M-15, SHV-OSBL, TEM-OSBL
		1	KPC-2, CTX-M-24, SHV-OSBL
		2	KPC-2, CTX-M-65, SHV-OSBL, TEM-OSBL
		1	KPC-2, CTX-M-67, SHV-OSBL
		1	KPC-2, CTX-M-90, SHV-OSBL, TEM-OSBL
		1	KPC-2, GES-6, SHV-OSBL, TEM-OSBL
		1	KPC-2, VEB-1, SHV-OSBL
		16	KPC-2, VEB-1, SHV-OSBL, TEM-OSBL
		1	KPC-3, SHV-12
		12	KPC-3, SHV-12, TEM-OSBL
		1	KPC-3, SHV-12, CTX-M-12, TEM-OSBL
		1	KPC-3, SHV-28, CTX-M-15, TEM-OSBL
		1	KPC-3, CTX-M-2, SHV-OSBL, TEM-OSBL
		1	KPC-3, CTX-M-15, SHV-OSBL
		9	KPC-3, CTX-M-15, SHV-OSBL, TEM-OSBL
		2	KPC-9, VEB-1, SHV-OSBL, TEM-OSBL
		1	KPC-12, SHV-2A
	R. ornithinolytica	1	KPC-3, SHV-5, TEM-OSBL
KPC + AmpC ± OSBL	C. amalonaticusc	1	KPC-2, TEM-OSBL
	C. farmeric	1	KPC-3
	C. freundiic	4	KPC-2
	C. koseric	1	KPC-2
	E. aerogenesc	1	KPC-2
		1	KPC-2, TEM-OSBL
	E. asburiaec	1	KPC-2
	E. cloacaec	1	KPC-2
		1	KPC-2, SHV-OSBL
		3	KPC-2, TEM-OSBL
	E. coli	1	KPC-2, CMY-2
		1	KPC-2, CMY-2, TEM-OSBL
	K. pneumoniae	1	KPC-2, ACT-type, SHV-OSBL
		1	KPC-2, CMY-2, SHV-OSBL
		1	KPC-2, CMY-2, SHV-OSBL, TEM-OSBL
		1	KPC-2, CMY-4, SHV-OSBL
		1	KPC-2, DHA-1, SHV-OSBL
	M. morganiic	1	KPC-2, TEM-OSBL
	S. marcescensc	2	KPC-2
	P. aeruginosac	28	KPC-2
KPC ± OSBL	E. coli	5	KPC-2
		6	KPC-2, TEM-OSBL
		2	KPC-3
		5	KPC-3, TEM-OSBL
		2	KPC-18, TEM-OSBL
	K. pneumoniae	10	KPC-2
		53	KPC-2, SHV-OSBL
		3	KPC-2, TEM-OSBL
		70	KPC-2, SHV-OSBL, TEM-OSBL
		2	KPC-3
		31	KPC-3, SHV-OSBL
		4	KPC-3, TEM-OSBL
		98	KPC-3, SHV-OSBL, TEM-OSBL

^aMBL, metallo-β-lactamase; ESBL, extended-spectrum β-lactamase; OSBL, original spectrum β-lactamase (includes TEM-1, TEM-2, SHV-1, and SHV-11).

^bPresumed to also carry the intrinsic chromosomally encoded ESBL common to this species.

^cPresumed to also carry the intrinsic chromosomally encoded AmpC β-lactamase common to this species.

The in vitro activities of β-lactam agents and comparators against the overall collection of Enterobacteriaceae, P. aeruginosa, and subsets of KPC-positive isolates coproducing additional β-lactamases from Ambler class A, B, C, and D were determined (Table 4). As expected, the activities of β-lactams, including aztreonam, ceftazidime, cefepime, meropenem, imipenem, and piperacillin-tazobactam, were greatly reduced against the overall subset of KPC-producing Enterobacteriaceae, with <5% of isolates susceptible to any of these agents. Combination of avibactam, a non-β-lactam β-lactamase inhibitor, with aztreonam or ceftazidime enhanced the activities of these β-lactams against KPC-positive isolates of Enterobacteriaceae at least 64-fold. Aztreonam-avibactam resulted in MIC₉₀ values of 0.5 to 1 μg/ml against all KPC-positive subsets, compared to MIC₉₀ values of >128 μg/ml for aztreonam. The MIC₉₀ values for ceftazidime-avibactam and ceftazidime were 2 to 4 μg/ml and >128 μg/ml, respectively, against KPC-positive isolates that did not coproduce a metallo-β-lactamase (MBL). The activities of agents from other drug classes against KPC-positive subsets were affected to different degrees; for example, the susceptibility to amikacin ranged from 42.5 to 78.6%, depending on the combination of coproduced β-lactamases, and the susceptibility to tigecycline ranged from 71.4 to 92.2%. The activity of colistin against subsets carrying different combinations of β-lactamases also varied, with susceptibilities ranging from 75.8 to 100%. Two isolates producing KPC and OXA-163 (OXA-48-like) showed low MIC values of ceftazidime-avibactam, aztreonam-avibactam, tigecycline, and colistin, but only aztreonam-avibactam and tigecycline were active in vitro against all six isolates carrying KPC and an MBL (Table 4).

TABLE 4.

In vitro activities of antimicrobial agents tested against carbapenem-nonsusceptible KPC-producing isolates collected in 2012 to 2014

Organism subset (n) and agenta	MIC (μg/ml)			% susceptibleb
	Range	50%	90%
All Enterobacteriaceae (38,266)
Ceftazidime	≤0.015 to >128	0.25	64	76.9
Ceftazidime-avibactamc	≤0.015 to >128	0.12	0.5	99.5
Aztreonam	≤0.015 to >128	0.12	64	75.7
Aztreonam-avibactamc	≤0.015 to >128	0.03	0.12	NA
Cefepime	≤0.12 to >16	≤0.12	>16	78.8
Meropenem	≤0.004 to >8	0.03	0.12	97.3
Imipenem	≤0.03 to >8	0.25	2	85.3
Piperacillin-tazobactam	≤0.25 to >128	2	64	84.7
Amikacin	≤0.25 to >32	2	8	96.6
Tigecycline	≤0.015 to >8	0.5	2	92.9
Colistind	≤0.12 to >4	≤0.12	>4	83.2
KPC-positive Enterobacteriaceae
All (557)
Ceftazidime	0.12 to >128	>128	>128	3.9
Ceftazidime-avibactam	≤0.015 to >128	1	4	97.5
Aztreonam	0.06 to >128	>128	>128	1.3
Aztreonam-avibactam	≤0.015 to 8	0.25	0.5	NA
Cefepime	≤0.12 to >16	>16	>16	4.8
Meropenem	0.06 to >8	>8	>8	3.1
Imipenem	0.5 to >8	>8	>8	0.5
Piperacillin-tazobactam	0.5 to >128	>128	>128	0.9
Amikacin	≤0.25 to >32	32	>32	48.3
Tigecycline	0.06 to 8	1	2	91.6
Colistin	≤0.12 to >4	0.03	>4	83.3
KPC ± OSBL (291)
Ceftazidime	1 to >128	128	>128	3.8
Ceftazidime-avibactam	≤0.015 to 128	1	4	98.6
Aztreonam	4 to >128	>128	>128	1.0
Aztreonam-avibactam	≤0.015 to 8	0.12	0.5	NA
Cefepime	≤0.12 to >16	>16	>16	5.5
Meropenem	0.25 to >8	>8	>8	3.1
Imipenem	2 to >8	>8	>8	0.0
Piperacillin-tazobactam	2 to >128	>128	>128	0.7
Amikacin	0.5 to >32	32	32	49.1
Tigecycline	0.06 to 8	1	2	91.8
Colistin	≤0.12 to >4	0.03	>4	89.0
KPC + ESBL ± OSBLe (219)
Ceftazidime	1 to >128	>128	>128	1.4
Ceftazidime-avibactam	≤0.015 to 16	1	4	99.1
Aztreonam	2 to >128	>128	>128	1.4
Aztreonam-avibactam	≤0.015 to 4	0.25	0.5	NA
Cefepime	≤0.12 to >16	>16	>16	2.3
Meropenem	0.5 to >8	>8	>8	1.4
Imipenem	0.5 to >8	>8	>8	0.9
Piperacillin-tazobactam	8 to >128	>128	>128	0.9
Amikacin	≤0.25 to >32	32	>32	42.5
Tigecycline	0.06 to 8	1	2	92.2
Colistin	≤0.12 to >4	0.03	>4	75.8
KPC + AmpC ± OSBLf (25)
Ceftazidime	0.12 to >128	32	>128	24.0
Ceftazidime-avibactam	0.03 to 2	0.5	2	100
Aztreonam	0.06 to >128	>128	>128	4.0
Aztreonam-avibactam	≤0.015 to 4	0.12	1	NA
Cefepime	≤0.12 to >16	16	>16	16.0
Meropenem	0.06 to >8	8	>8	16.0
Imipenem	2 to >8	8	>8	0.0
Piperacillin-tazobactam	0.5 to >128	>128	>128	4.0
Amikacin	0.5 to >32	4	>32	76.0
Tigecycline	0.06 to 4	1	2	92.0
Colistin	≤0.12 to >4	≤0.12	>4	80.0
KPC + ESBL + AmpC ± OSBLf (14)
Ceftazidime	2 to >128	64	>128	14.3
Ceftazidime-avibactam	0.25 to 4	1	2	100
Aztreonam	16 to >128	>128	>128	0.0
Aztreonam-avibactam	0.03 to 1	0.25	0.5	NA
Cefepime	≤0.12 to >16	>16	>16	14.3
Meropenem	0.5 to >8	4	>8	7.1
Imipenem	1 to >8	8	>8	7.1
Piperacillin-tazobactam	64 to >128	>128	>128	0.0
Amikacin	≤0.25 to >32	4	32	78.6
Tigecycline	0.25 to 8	1	4	71.4
Colistin	≤0.12 to 0.06	≤0.12	0.06	100
KPC + OXA-48-like + OSBL ± ESBL (2)
Ceftazidime	128 to >128			0.0
Ceftazidime-avibactam	1 to 2			100
Aztreonam	64 to >128			0.0
Aztreonam-avibactam	0.25 to 0.5			NA
Cefepime	>16 to >16			0.0
Meropenem	2 to >8			0.0
Imipenem	4 to 8			0.0
Piperacillin-tazobactam	>128 to >128			0.0
Amikacin	16 to >32			50.0
Tigecycline	2 −2			100
Colistin	≤0.12 to ≤0.12			100
KPC + MBL ± ESBL ± AmpC ± OSBL (6)
Ceftazidime	>128 to >128			0.0
Ceftazidime-avibactam	>128 to >128			0.0
Aztreonam	>128 to >128			0.0
Aztreonam-avibactam	0.5 to 1			NA
Cefepime	>16 to >16			0.0
Meropenem	>8 to >8			0.0
Imipenem	>8 to >8			0.0
Piperacillin-tazobactam	>128 to >128			0.0
Amikacin	16 to >32			16.7
Tigecycline	0.5 to 2			100
Colistin	≤0.12 to >4			50.0
All P. aeruginosa (8,010)
Ceftazidime	0.06 to >128	2	64	77.4
Ceftazidime-avibactam	0.06 to >128	2	8	92.4
Aztreonam	≤0.015 to >128	8	32	61.4
Aztreonam-avibactam	≤0.015 to >128	8	32	NA
Cefepime	≤0.12 to >16	4	16	78.6
Meropenem	≤0.06 to >8	0.5	>8	73.3
Imipenem	≤0.03 to >8	2	>8	61.7
Piperacillin-tazobactam	≤0.25 to >128	8	>128	69.1
Amikacin	≤0.25 to >32	4	16	90.2
Colistin	≤0.12 to >8	0.5	1	99.5
KPC-positive P. aeruginosa
All (29)
Ceftazidime	64 to >128	64	>128	0.0
Ceftazidime-avibactam	4 to 64	8	32	75.9
Aztreonam	>128 to >128	>128	>128	0.0
Aztreonam-avibactam	8 to >128	32	64	NA
Cefepime	0.5 to >16	>16	>16	3.4
Meropenem	>8 to >8	>8	>8	0.0
Imipenem	>8 to >8	>8	>8	0.0
Piperacillin-tazobactam	>128 to >128	>128	>128	0.0
Amikacin	1 to >32	8	>32	75.9
Colistin	≤0.06 to >8	0.5	2	96.6
KPC + AmpCf (28)
Ceftazidime	64 to >128	64	>128	0.0
Ceftazidime-avibactam	4 to 64	8	32	78.6
Aztreonam	>128 to >128	>128	>128	0.0
Aztreonam-avibactam	8 to >128	32	64	NA
Cefepime	0.5 to >16	>16	>16	3.6
Meropenem	>8 to >8	>8	>8	0.0
Imipenem	>8 to >8	>8	>8	0.0
Piperacillin-tazobactam	>128 to >128	>128	>128	0.0
Amikacin	1 to >32	8	>32	78.6
Colistin	≤0.06 to >8	0.5	2	96.4
KPC + AmpC + MBLf (1)
Ceftazidime	64			0.0
Ceftazidime-avibactam	64			0.0
Aztreonam	>128			0.0
Aztreonam-avibactam	16			NA
Cefepime	>16			0.0
Meropenem	>8			0.0
Imipenem	>8			0.0
Piperacillin-tazobactam	>128			0.0
Amikacin	>32			0.0
Colistin	2			100

^aMBL, metallo-β-lactamase; ESBL, extended-spectrum β-lactamase; OSBL, original-spectrum β-lactamase (includes TEM-1, TEM-2, SHV-1, and SHV-11).

^bSusceptibility percentages were determined using CLSI interpretive criteria. FDA breakpoints were applied for ceftazidime-avibactam (≤8 μg/ml, susceptible; ≥16 μg/ml, resistant) and tigecycline (≤2 μg/ml, susceptible; 4 μg/ml, intermediate; ≥8 μg/ml, resistant). EUCAST breakpoints were applied for colistin tested against Enterobacteriaceae (≤2 μg/ml, susceptible; ≥4 μg/ml, resistant).

^cAztreonam-avibactam and ceftazidime-avibactam were tested at a fixed concentration of 4 μg/ml avibactam.

^dColistin was tested with 0.002% polysorbate 80.

^eIncludes the presumed chromosomally encoded ESBL common to K. oxytoca.

^fIncludes plasmid-encoded and presumed chromosomally encoded AmpC β-lactamases common to Enterobacter spp., Citrobacter spp., M. morganii, S. marcescens, and P. aeruginosa.

The majority of antimicrobial agents tested were inactive against P. aeruginosa isolates producing KPC (susceptibilities of <4%), but 76% of the overall subset were susceptible to ceftazidime-avibactam or amikacin, and 96.6% were susceptible to colistin. However, only colistin remained active against the one P. aeruginosa isolate that coproduced KPC and an MBL (MIC, 2 μg/ml) (Table 4).

DISCUSSION

KPC carbapenemases hydrolyze penicillins, oxyimino-cephalosporins, cephamycins, monobactams, and carbapenems as well as the commercially available β-lactamase inhibitors clavulanic acid, sulbactam, and tazobactam (12, 24). Carbapenem MIC values against KPC-producing bacteria can range from susceptible to fully resistant, with elevated KPC production due to increased bla_KPC copy number and/or deletions in the upstream promoter region associated with higher MIC values in some isolates (12, 25, 26). Production of KPC is often accompanied by loss of either or both of the OmpK35 and OmpK36 porins, which further decreases susceptibility to carbapenems (25, 27–29). Four isolates (one K. pneumoniae and two E. coli collected from the same medical center in Colombia and one K. pneumoniae collected in the United States) that were susceptible to all tested carbapenems were identified, and they were presumed not to express KPC at significant levels.

bla_KPC has disseminated from K. pneumoniae to P. aeruginosa and multiple species of Enterobacteriaceae, including E. coli, K. oxytoca, Enterobacter spp., Citrobacter spp., S. marcescens, M. morganii, and R. ornithinolytica, as described in this study and by others (12, 30). bla_KPC has also been reported in Acinetobacter baumannii, Proteus mirabilis, Providencia stuartii, Pantoea agglomerans, Leclercia adecarboxylata, Kluyvera spp., Pseudomonas putida, and Salmonella spp. (12, 30–33). Intra- and interspecies spread of bla_KPC is attributed to transposition of Tn4401, an active transposon with no target site specificity, to a variety of broad- and narrow-host-range plasmids capable of conjugation (34). Mathers et al. described three bla_KPC-bearing plasmids identified during a hospital outbreak; one highly mobile plasmid was found in 11 isolates comprised of 9 unique strains (3 K. pneumoniae, 4 E. cloacae, and 1 each E. asburiae and C. freundii) collected from patients in various hospital units during an 8-month period, whereas the other two plasmids were found in 2 K. oxytoca isolates and 1 E. coli isolate, respectively. It should be noted that only approximately one-third of the affected patients had received treatment with a carbapenem, and one patient harbored two isolates (K. pneumoniae and E. asburiae) carrying the same KPC-encoding plasmid (14). In another study, three different KPC-producing species were sequentially collected from a patient over a 5-month period. Molecular analyses indicated that bla_KPC was first transferred between plasmids carried by K. pneumoniae and E. coli via a Tn4401-mediated event, followed by conjugation of the bla_KPC-bearing plasmid from E. coli into S. marcescens (35). The rapid and global spread of bla_KPC has also been facilitated by carriage by K. pneumoniae strains belonging to clonal complex 258 (CC258), most frequently ST258 (36, 37). CC258 isolates tend to be multidrug resistant (MDR). In addition to bla_KPC-bearing plasmids conferring resistance to β-lactams, members of CC258 often carry additional plasmids encoding resistance to aminoglycosides, trimethoprim, sulfonamides, and macrolides (16, 36, 38, 39). CC258 isolates also possess chromosomal mutations in gyrA and parC conferring fluoroquinolone resistance, and colistin-resistant ST258 isolates have been reported (36, 37, 39).

Treatment options available for managing patients infected with carbapenem-resistant or MDR pathogens have not kept pace with the emergence of resistance mechanisms in the patient population. Patients hospitalized in long-term and acute-care facilities are at significant risk for acquiring isolates producing KPC (40, 41). KPC-producing MDR isolates often remain susceptible only to tigecycline, polymyxins, and some aminoglycosides (e.g., gentamicin or amikacin); however, monotherapy with tigecycline or colistin is frequently associated with high treatment failure rates (4, 8–10, 42). Ceftazidime-avibactam was recently used in combination with ertapenem to successfully treat a patient infected with a KPC-producing K. pneumoniae isolate that had become resistant to tigecycline and colistin after treatment for successive nosocomial infections (43). However, one KPC-producing isolate that was resistant to ceftazidime-avibactam via an unknown mechanism has also been reported (44). In this study, avibactam restored the in vitro activity of both ceftazidime and aztreonam against KPC-producing isolates of Enterobacteriaceae, including, in the case of aztreonam-avibactam, activity against isolates that coproduced MBLs.

Reports of the emergence of colistin-resistant KPC-producing K. pneumoniae potentially further limit the number of therapeutic options available to treat infections caused by these challenging pathogens. Ceftazidime-avibactam and aztreonam-avibactam demonstrate potent in vitro activity against KPC-producing Enterobacteriaceae and may be powerful additions to the existing armamentarium of antimicrobial agents.

Funding Statement

This investigation was funded by AstraZeneca Pharmaceuticals as part of a global surveillance program. The sponsor approved the overall study design. All investigative sites were recruited and study supplies were provided by IHMA, Inc. Analysis of the final MIC and molecular data was performed by IHMA, Inc., and was independent of sponsor analysis for this study.