Assessment of the In Vitro Activity of Ceftazidime-Avibactam against Multidrug-Resistant Klebsiella spp. Collected in the INFORM Global Surveillance Study, 2012 to 2014

Abstract

Increasing resistance in Gram-negative bacilli, including Klebsiella spp., has reduced the utility of broad-spectrum cephalosporins. Avibactam, a novel non-β-lactam β-lactamase inhibitor, protects β-lactams from hydrolysis by Gram-negative bacteria that produce extended-spectrum β-lactamases (ESBLs) and serine carbapenemases, including Ambler class A and/or class C and some class D enzymes. In this analysis, we report the in vitro activity of ceftazidime-avibactam and comparators against multidrug-resistant (MDR) Klebsiella spp. from the 2012-2014 INFORM surveillance study. Isolates collected from 176 sites were sent to a central laboratory for confirmatory identification and tested for susceptibility to ceftazidime-avibactam and comparator agents, including ceftazidime alone. A total of 2,821 of 10,998 (25.7%) Klebsiella species isolates were classified as MDR, based on resistance to three or more classes of antimicrobials. Among the MDR isolates, 99.4% had an ESBL screen-positive phenotype, and 27.4% were not susceptible to meropenem as an example of a carbapenem. Ceftazidime-avibactam was highly active against MDR isolates, including ESBL-positive and serine carbapenemase-producing isolates, with MIC_50/90 values of 0.5/2 μg/ml and 96.6% of all MDR isolates and ESBL-positive MDR isolates inhibited at the FDA breakpoint (MIC value of ≤8 μg/ml). Ceftazidime-avibactam did not inhibit isolates producing class B enzymes (metallo-β-lactamases) either alone or in combination with other enzymes. These in vitro results support the continued investigation of ceftazidime-avibactam for the treatment of MDR Klebsiella species infections.

INTRODUCTION

Cephalosporins have provided essential treatment options for numerous infection types since their discovery, with many having a broad spectrum of activity against both Gram-positive and -negative pathogens. Increasing multidrug resistance (MDR) in Gram-negative bacteria has reduced the utility of broad-spectrum cephalosporins, including ceftazidime (1). The prevalence, mobility, and multiplicity of β-lactamases, including extended-spectrum β-lactamases (ESBLs), AmpC cephalosporinases, and carbapenemases, are the leading causes of cephalosporin resistance associated with patient infections (2). Although ceftazidime is active against many Gram-negative bacteria, it is inactive against many strains that produce ESBLs and/or highly expressed class C β-lactamases.

As resistance to broad-spectrum cephalosporins has increased, carbapenems have become the main β-lactam choice for therapy; however, the increased use of carbapenems has in turn led to the appearance of carbapenem-resistant strains, including those harboring genes encoding KPC or OXA-48 enzymes (3). A high proportion of MDR isolates are nonsusceptible to carbapenems, and carbapenem-resistant Enterobacteriaceae are commonly cross-resistant to other classes of antibiotics, such as fluoroquinolones, trimethoprim-sulfamethoxazole, and aminoglycosides, resulting in limited therapeutic options to treat infections caused by these pathogens (4–7).

Avibactam is a novel non-β-lactam β-lactamase inhibitor that has been developed for use in combination with the β-lactam antibiotic ceftazidime and is approved in the United States by the FDA for the treatment of adults with complicated intra-abdominal infections, in combination with metronidazole, and complicated urinary tract infections, including kidney infections (pyelonephritis), who have limited or no alternative treatment options. Unlike earlier β-lactamase inhibitors, which were developed for use against Ambler class A enzymes (8), avibactam displays a broader spectrum of activity. Specifically, avibactam is capable of protecting β-lactams from hydrolysis by ESBLs and carbapenemases, including Ambler class A and/or class C enzyme producers and some class D enzymes (9–11). Avibactam has been evaluated in combination with ceftazidime to assess whether it enhances ceftazidime's spectrum of antibacterial activity to include Gram-negative bacteria that produce ESBLs and/or carbapenemases, including KPC-producing strains.

Klebsiella pneumoniae and Klebsiella oxytoca have long been recognized as significant human pathogens, causing community-associated pneumonia, nosocomial infections in both adult and pediatric populations, and neonatal sepsis (12). In recent years, the ability of Klebsiella spp. to become resistant to antimicrobial agents, including carbapenems and colistin, has reduced therapeutic options dramatically (4). Mechanisms for resistance in Klebsiella spp. are diverse and include the production of carbapenemases, changes in outer membrane proteins (porin loss), and the upregulation of efflux systems (13). Dissemination of KPC-producing K. pneumoniae is largely attributed to the expansion of a single dominant strain, ST258 (14).

In this analysis, we report on the in vitro antibacterial activity of ceftazidime-avibactam compared with those of ceftazidime alone and nine additional comparators against MDR Klebsiella spp. collected during the 2012-2014 INFORM (International Network For Optimal Resistance Monitoring) global surveillance study.

(Some of these data were presented previously in abstract form [15].)

MATERIALS AND METHODS

In 2012 to 2014, a total of 10,998 Klebsiella isolates (9,098 K. pneumoniae isolates and 1,900 K. oxytoca isolates) were collected from 176 sites in the INFORM surveillance program with 39 participating countries in Asia/Pacific, Europe, Latin America, and Africa/Middle East. In 2012 to 2013, each site collected and identified consecutive fresh clinical isolates from documented intra-abdominal infections (IAI), urinary tract infections (UTI), skin and soft tissue infections (SSTI), and lower respiratory tract infections (LRTI). In 2014, the acceptable sources were expanded to include blood infections. Only one isolate of the same species per patient was included in the surveillance study. The distribution of Klebsiella spp. collected was 3,263 (29.7%) from LRTI, 3,292 (29.9%) from UTI, 2,175 (19.8%) from SSTI, 1,821 (16.6%) from IAI, 413 (3.8%) from blood, and 34 (0.3%) from unknown sources. All isolates were sent to a central laboratory, International Health Management Associates, Inc. (IHMA) (Schaumburg, IL, USA), where the isolates were further evaluated and stored. Isolates were identified at each participating center and confirmed at IHMA using matrix-assisted laser desorption ionization−time of flight (MALDI-TOF) mass spectrometry (Bruker Daltonics, Bremen, Germany).

The isolates were tested for susceptibility to multiple drugs, and the results for ceftazidime, ceftazidime-avibactam (avibactam at a fixed concentration of 4 μg/ml), amikacin, cefepime, colistin (with 0.002% polysorbate 80 [P-80]), doripenem, ertapenem, imipenem, meropenem, piperacillin-tazobactam, and tigecycline are presented in this study. Ertapenem was tested only in 2012 to 2013. The MICs were determined using broth microdilution panels prepared at IHMA. All broth microdilution testing, including panel manufacture, inoculation, incubation, and interpretation, was conducted following the Clinical and Laboratory Standards Institute (CLSI) M07-A9 and M100-S25 guidelines (16, 17). Food and Drug Administration (FDA) breakpoint criteria were applied for tigecycline (18) and ceftazidime-avibactam (19), and European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoint criteria for colistin tested without P-80 were applied for colistin (20, 21). Isolates were screened for ESBL activity according to the CLSI guidelines using ceftazidime and/or aztreonam, followed by confirmation by the clavulanate inhibition test (17). MDR was defined by resistance according to the CLSI guidelines (17) against three or more drug classes, including aminoglycosides (amikacin), β-lactam/β-lactamase inhibitor combinations (amoxicillin-clavulanate, piperacillin-tazobactam), monobactams (aztreonam), cephalosporins (cefepime, ceftazidime), carbapenems (doripenem, ertapenem, imipenem, meropenem), and fluoroquinolones (levofloxacin). Klebsiella spp. are intrinsically resistant against penicillins, and therefore this class of antibacterials was not included as a criterion for MDR. Quality control testing using Escherichia coli ATCC 25922 and 35218, Pseudomonas aeruginosa ATCC 27853, and K. pneumoniae ATCC 700603 was performed on each day of testing following the CLSI guidelines (17).

Isolates that were positive for ESBL activity based on the phenotypic confirmatory test or nonsusceptible to one or more of the tested carbapenems were analyzed for genes encoding ESBLs (TEM, SHV, VEB, CTX-M-type, PER, GES), carbapenemases (KPC, NDM, IMP, VIM, SPM, OXA-48-type), and AmpC β-lactamases (CMY, DHA, FOX, MOX, ACC, MIR, ACT). Genes were detected using a combination of microarray (Check-MDR CT101; Check-Points B.V., Wageningen, Netherlands) and multiplex PCR assays as described previously (22). Detected genes encoding ESBLs, carbapenemases, and AmpC enzymes were sequenced and compared to public databases available from the National Center for Biotechnology Information (www.ncbi.nlm.nih.gov) and the Lahey Clinic (www.lahey.org/studies/).

RESULTS

Of 10,998 Klebsiella isolates collected, 2,821 (25.7%) were MDR, consisting of 2,739/9,098 K. pneumoniae isolates (30.1%) and 82/1,900 K. oxytoca isolates (4.3%). MDR isolates were found across all indications and across the globe. The distribution of MDR Klebsiella isolates in this analysis was 899 from LRTI (27.6% of LRTI isolates), 827 from UTI (25.1% of UTI isolates), 608 SSTI isolates (28.0% of SSTI isolates), 355 from IAI (19.5% of IAI isolates), 122 from blood (29.5% of blood isolates), and 10 from unknown sources (29.4% of such isolates). The distribution by country is shown in Table 1. There was a wide range of percentages of MDR isolates across the different countries, with some countries having a relatively large proportion of MDR isolates (Brazil, Nigeria, Russia [>50%]) and others a small proportion (Australia, Denmark, the Netherlands, Sweden, United Kingdom [<5%]).

TABLE 1.

Distribution of Klebsiella spp. and percentages of multidrug resistance by country

Country or region	Klebsiella spp., all (no.)a	% of global total	Klebsiella spp., MDR (no.)b	% of MDR
Argentina	301	2.7	121	40.2
Australia	323	2.9	10	3.1
Austria	226	2.1	22	9.7
Belgium	372	3.4	43	11.6
Brazil	238	2.2	126	52.9
Chile	283	2.6	127	44.9
Chinac	367	3.3	74	20.2
Colombia	179	1.6	48	26.8
Czech Republic	288	2.6	70	24.3
Denmark	206	1.9	6	2.9
France	430	3.9	60	14.0
Germany	415	3.8	41	9.9
Greece	360	3.3	161	44.7
Hong Kongd	30	0.3	3	10.0
Hungary	243	2.2	71	29.2
Ireland	20	0.2	5	25.0
Israel	380	3.5	105	27.6
Italy	427	3.9	160	37.5
Japan	179	1.6	9	5.0
Kenya	86	0.8	29	33.7
Korea, South	266	2.4	69	25.9
Kuwait	230	2.1	36	15.7
Malaysia	163	1.5	36	22.1
Mexico	406	3.7	84	20.7
Netherlands	194	1.8	6	3.1
Nigeria	127	1.2	64	50.4
Philippines	322	2.9	103	32.0
Poland	215	2.0	86	40.0
Portugal	412	3.7	136	33.0
Romania	237	2.2	100	42.2
Russia	762	6.9	442	58.0
South Africa	290	2.6	62	21.4
Spain	456	4.1	54	11.8
Sweden	212	1.9	0	0.0
Taiwan	295	2.7	57	19.3
Thailand	287	2.6	76	26.5
Turkey	275	2.5	47	17.1
United Kingdom	222	2.0	7	3.2
Venezuela	274	2.5	65	23.7
Global	10,998	100	2,821	25.7

^aA total of 9,098 K. pneumoniae isolates and 1,900 K. oxytoca isolates were studied.

^bMultidrug resistance (MDR) is defined by resistance to three or more drug classes, including classes not reported in this study.

^cIsolates collected from mainland China and Hong Kong in 2012 to 2013. No isolates were obtained from patients in mainland China in 2014 due to export restrictions.

^dIsolates collected from Hong Kong in 2014.

Based on the MIC₉₀ values, ceftazidime-avibactam was one of the most active agents tested against isolates of MDR Klebsiella spp., together with colistin and tigecycline, with MIC_50/90 values of 0.5/2 μg/ml, ≤0.12/0.25 μg/ml, and 1/2 μg/ml, respectively (Table 2). The addition of avibactam to ceftazidime provided a ≥128-fold reduction in MIC_50/90 values compared to those for ceftazidime alone. Globally, 96.6% of the MDR isolates of Klebsiella spp. exhibited ceftazidime-avibactam MICs of ≤8 μg/ml and can be considered susceptible according to the FDA breakpoint. Tigecycline and colistin were the only other agents tested to which >90% of MDR isolates were susceptible. Carbapenem susceptibility for all MDR isolates ranged from 59.6% for ertapenem to 74.9% for doripenem. Susceptibility to amikacin was slightly higher (at 78.1%) among these MDR Klebsiella spp. (Table 2). Similar results were observed for MDR ESBL screen-positive isolates (not shown).

TABLE 2.

In vitro activity of ceftazidime-avibactam and comparator agents against multidrug-resistant Klebsiella spp.

Antimicrobial tested in MDR Klebsiella spp. (2,821)	MIC (μg/ml)			% of isolates:
	MIC50	MIC90	Range	Susceptible	Intermediate	Resistant
Ceftazidime-avibactama	0.5	2	≤0.015–>128	96.6		3.4
Ceftazidime	128	>128	0.12–>128	2.6	1.6	95.8
Amikacin	4	>32	≤0.25–>32	78.1	9.6	12.3
Cefepimeb	>16	>16	≤0.12–>16	6.0	9.4	84.6
Colistinc	≤0.12	0.25	≤0.12–>4	93.1		6.9
Doripenem	0.12	>4	0.015–>4	74.9	3.6	21.5
Ertapenemd	0.5	>1	≤0.002–>1	59.6	7.2	33.2
Imipenem	0.25	>8	0.06–>8	73.2	3.6	23.2
Meropenem	0.06	>8	0.008–>8	72.6	3.9	23.5
Piperacillin-tazobactam	>128	>128	0.5–>128	22.9	18.3	58.8
Tigecyclinea	1	2	0.03−8	93.1	5.7	1.2

^aPercent susceptible based on FDA breakpoints (18, 19).

^bFor cefepime, the intermediate category was replaced by susceptible-dose dependent (SDD) in 2014.

^cColistin was tested in the presence of 0.002% polysorbate-80 (21); percent susceptible was based on the EUCAST breakpoints for colistin (20).

^dErtapenem values are based on data from 2012 to 2013 only (n = 1,566); ertapenem was not tested in 2014.

Among the meropenem-nonsusceptible subgroup of MDR isolates (773 strains, 27.4% of MDR isolates), 88.1% were susceptible to ceftazidime-avibactam based on the FDA interpretive criterion of susceptibility (a MIC of ≤8 μg/ml) (Table 3), while 96.5% and 96.6%, respectively, of the ceftazidime-nonsusceptible and ESBL screen-positive subsets were susceptible by the same criterion (Table 3). Of the meropenem-nonsusceptible isolates, 10.2% were not inhibited by ≤8 μg/ml of ceftazidime-avibactam due to the production of metallo-β-lactamases.

TABLE 3.

Frequencies and cumulative percentages of multidrug-resistant Klebsiella spp. and specific resistance phenotypes inhibited by ceftazidime and ceftazidime-avibactam^a

^aMDR, multidrug resistance (defined by resistance to three or more drug classes, including classes not included in this analysis); no., number of isolates; Cum%, cumulative percentage inhibited; MEM, meropenem; CAZ, ceftazidime; CAZ-AVI, ceftazidime with 4 μg/ml avibactam; NS, nonsusceptible (MIC ≥2 μg/ml for MEM and MIC ≥8 μg/ml for CAZ); ESBL positive, presence of extended-spectrum-β-lactamase activity based on a CLSI screen test result (aztreonam and/or ceftazidime MIC values of ≥2 μg/ml) (17). MIC₉₀s are shaded. The FDA susceptibility breakpoint for ceftazidime-avibactam is MIC ≤8 μg/ml (19).

Table 4 shows the MIC₉₀ values and percentages of susceptible MDR isolates by region. Regionally, the MIC₉₀ values for ceftazidime-avibactam were consistent at 2 μg/ml for isolates from Africa/Middle East, Europe, and Latin America and were one doubling dilution higher in Asia/Pacific isolates at 4 μg/ml. In all regions, the addition of avibactam increased the activity of ceftazidime, with a ≥64-fold decrease in MIC₉₀ values compared to those for ceftazidime alone. The in vitro activities of the other antimicrobials tested varied by geographic region (Table 4). The susceptibility to tigecycline and colistin was ≥91% in all regions. The susceptibility of the MDR isolates to amikacin ranged from 76.4% in Latin America to 82.4% in Africa/Middle East. Doripenem, imipenem, and meropenem were active against MDR isolates from Africa/Middle East (83 to 85% susceptible) and Asia/Pacific (82 to 87% susceptible), while ertapenem showed reduced activity in these regions with 67% of Asia/Pacific and 75% of Africa/Middle East isolates susceptible. Lower susceptibilities were observed for all carbapenems in Europe (57 to 73% of isolates were susceptible) and Latin America (52 to 70% susceptible) than for other regions. All regions had cefepime susceptibility percentages ranging from 3 to 12%.

TABLE 4.

In vitro activity of ceftazidime-avibactam and comparators against MDR Klebsiella spp. by region

^a Percent susceptible at >90% values are shaded. AMK, amikacin; FEP, cefepime; CAZ, ceftazidime; CAZ-AVI, ceftazidime with 4 μg/ml avibactam; CST, colistin; DOR, doripenem; ETP, ertapenem; IPM, imipenem; MEM, meropenem; TZP, piperacillin-tazobactam; TGC, tigecycline.

^b Percent susceptible is based on the FDA breakpoint (18, 19).

^c Colistin was tested in the presence of 0.002% polysorbate-80 (21). Percent susceptible is based on EUCAST breakpoint for colistin (20).

Table 5 shows the cumulative frequency distribution of ceftazidime and ceftazidime-avibactam MICs against MDR Klebsiella spp. separated by Ambler class enzyme designations (36). For the 2,374 isolates encoding only a class A enzyme, including 426 encoding KPC, the addition of avibactam reduced the ceftazidime MIC₉₀ from >128 μg/ml to 2 μg/ml. Isolates encoding a class A enzyme in combination with a class C or a class D enzyme also exhibited a >64-fold reduction in MIC₉₀. Avibactam decreased the ceftazidime MIC values at least 64-fold for isolates encoding a class C enzyme alone. A total of 81 MDR isolates produced Ambler class B enzymes, including NDM, IMP, and VIM, either alone or in combination with a class A, C, or D enzyme. Isolates encoding a class B enzyme showed no meaningful reductions in ceftazidime MIC values with the addition of avibactam. Avibactam addition restored the activity of ceftazidime against eight isolates in which no β-lactamase was detected; it is likely that these isolates carry undetected class A or class C enzymes.

TABLE 5.

MICs and cumulative percent inhibited distributions for ceftazidime and ceftazidime-avibactam against multidrug-resistant Klebsiella isolates producing Ambler class A, B, C, or D enzymes^a

^aMDR, multidrug resistance (defined by resistance to three or more drug classes, including classes not included in this analysis); no., number of isolates; Cum%, cumulative percentage inhibited; CAZ, ceftazidime; CAZ-AVI, ceftazidime with 4 μg/ml avibactam. MIC₉₀ values (shaded) are only reported for >10 isolates. The FDA susceptibility breakpoint for ceftazidime-avibactam is MIC ≤8 μg/ml (19).

^b Class A: presence of bla_SHV, bla_TEM, bla_CTX-M, bla_VEB, bla_PER, bla_GES, and/or bla_KPC confirmed by PCR. K. oxytoca isolates are presumed to carry the intrinsic chromosomally encoded bla_OXY common to this species, though this was not confirmed by molecular methods. Class B: presence of bla_VIM, bla_IMP, and/or bla_NDM confirmed by PCR. Class C: presence of bla_ACC, bla_CMY, bla_DHA, bla_FOX, bla_MOX, and/or bla_ACT confirmed by PCR. Class D: presence of bla_OXA-48-like confirmed by PCR. No β-lactamase identified: isolates in which none of the tested β-lactamase genes were detected by PCR. Does not include isolates of K. oxytoca, which are presumed to carry the intrinsic chromosomally encoded bla_OXY.

There were 95 MDR Klebsiella species isolates overall with ceftazidime-avibactam MIC values of >8 μg/ml, collected from Europe (39 isolates), Asia/Pacific (33 isolates), Africa/Middle East (20 isolates), and Latin America (3 isolates). Seventy-nine isolates (83.2%; 72 K. pneumoniae isolates and 7 K. oxytoca isolates) carried genes encoding class B enzymes either alone or in combination with other β-lactamases. The remaining 16 isolates all produced class A enzymes (ESBLs or original-spectrum SHV- or TEM-type β-lactamases [OSBLs]). Of these, two K. pneumoniae isolates carrying only an SHV-type OSBL that tested with ceftazidime MIC values of >128 μg/ml were resistant to aztreonam, carbapenems, and piperacillin-tazobactam and were collected in China and Italy. Of the remaining 14 isolates, 5 isolates carried KPC (2 KPC-2 and 3 KPC-3), 1 isolate carried KPC-2 and an SHV-type ESBL, 6 isolates carried one or more ESBLs (SHV- and/or CTX-M-type), 1 isolate carried two class C β-lactamases (CMY-2 and DHA-1), and 1 isolate carried a class D β-lactamase (OXA-48, K. pneumoniae).

DISCUSSION

Cephalosporins continue to be important agents in the treatment of Gram-negative bacterial infections. However, their efficacy has been challenged by the increased prevalence of ESBLs, carbapenemases, and metallo-β-lactamases (1, 2, 23). MDR phenotypes are often associated with these enzymes, as well as altered porins, efflux pumps, and combinations of more than one enzyme type or other resistance mechanism. Carbapenems, often the first drug of choice against ESBL-producing pathogens (24), are now becoming less reliable. Carbapenem-resistant Enterobacteriaceae are increasing in prevalence in some regions and continue to spread globally (25, 26).

Avibactam is capable of protecting β-lactams from hydrolysis in Gram-negative bacteria that produce both ESBLs and carbapenemases, including Ambler class A and/or class C enzyme producers and some class D enzymes (9). In this analysis, ceftazidime-avibactam demonstrated potent in vitro activity against MDR Klebsiella spp. from a contemporary global population, including isolates producing ESBLs and serine carbapenemases, either alone or in combination with Ambler class C and D enzymes. Notably, 88.1% of meropenem-nonsusceptible MDR isolates retained susceptibility to ceftazidime-avibactam, an in vitro observation consistent with the combination being a potentially useful alternative therapeutic option to a carbapenem. Globally, 96.6% of MDR isolates were inhibited at the FDA susceptibility breakpoint (MIC of ≤8 μg/ml), with MIC_50/90 values reduced ≥128-fold compared to those for ceftazidime alone, going from 128/>128 μg/ml for ceftazidime to 0.5/2 μg/ml with the addition of avibactam, with no regional differences in susceptibility identified. It should be noted that India and Pakistan, in which some class B enzymes (metallo-β-lactamases) are considered endemic, were not included in this study (27). Ceftazidime-avibactam did not inhibit isolates producing class B enzymes either alone or in combination with other enzymes. In addition, a small number of isolates with ceftazidime-avibactam MIC values of >8 μg/ml that appeared to lack the transferable metallo-β-lactamases screened as part of this study were identified, indicating the existence of yet undetermined antimicrobial resistance mechanisms against ceftazidime-avibactam. These mechanisms might include modifications of β-lactamase or penicillin-binding protein (PBP) sequences or changes in drug efflux levels (28–31). Recently, a KPC-3-producing MDR K. pneumoniae isolate with a ceftazidime-avibactam MIC value of 32 μg/ml was reported by others (32).

The dissemination of β-lactamases has resulted in not only organisms that have increased resistance due to individual enzymes but also isolates that coproduce multiple enzymes. K. pneumoniae has become one such problematic pathogen, with many MDR strains impervious to last-in-line treatment options; new therapeutic alternatives are needed for this and other enteric pathogens (1, 4, 33–35). In our study, 15.0% (422) of the MDR Klebsiella isolates were found to carry a multiplicity of enzymes from different Ambler classes, with the ability to hydrolyze more than one class of antimicrobial agent. The in vitro results presented here demonstrate that ceftazidime-avibactam may be a valuable option for the treatment of infections caused by this increasingly difficult-to-treat pathogen.

Funding Statement

This investigation was funded by AstraZeneca Pharmaceuticals as part of the sponsored INFORM 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.