Activity of imipenem/relebactam on Klebsiella pneumoniae with different mechanisms of imipenem non-susceptibility.

BACKGROUND AND OBJECTIVES: Imipenem/relebactam (IMP/R) is a newly FDA approved β-lactam/β-lactamase inhibitor combination. Relebactam ability to restore IMP activity could differ according to the cause of imipenem non-susceptibility. Therefore, we investigated the in-vitro activity of IMP/R against Klebsiella pneumoniae with different mechanisms of imipenem non-susceptibility.
MATERIALS AND METHODS: Imipenem-nonsusceptible (IMP-NS) K. pneumoniae isolates were collected and characterized for β-lactamase encoding genes by multiplex PCR. For IMP-NS carbapenemase-negative isolates, study of Ompk35 & Ompk36 gene expression was performed by reverse transcription-PCR while efflux pump activity was studied by minimum inhibitory concentration (MIC) reduction assay using efflux pump inhibitor. Susceptibility testing of K. pneumoniae to IMP and IMP/R were achieved by broth microdilution (BMD) method.
RESULTS: During the study period, 140 isolates of IMP-NS K. pneumoniae were collected. BMD method showed that relebactam restored IMP susceptibility in 100%, 60% and 49% of isolates that only harbor AmpC, extended spectrum beta lactamase (ESBL) and carbapenemases, respectively. IMP/R was most potent against all bla KPC and 50% of bla OXA-48 _producing isolates. No demonstrable activity of IMP/R against K. pneumoniae harboring metallo-β-lactamases (MBLs). Out of 18 isolates with IMP non-suceptibility due to porins loss with overproduction of ESBL and/or AmpC, 14 (77.7%) isolates were IMP/R susceptible. IMP/R showed no activity against isolates with only efflux pump hyperactivity.
CONCLUSION: Relebactam could restore IPM activity in KPC or AmpC-producing IMP/NS K. pneumoniae but with no activity against MBL- producing isolates. Relebactam activity against isolates harbouring-bla OXA-48 or with altered Ompk35 & Ompk36 gene expression and efflux pump hyperactivity need further studies. Therefore, using IMP/R antibiotic in the treatment of infections caused by IMP/NS K. pneumoniae should be based on its molecular profile of IMP resistance to optimize the utility of IMP/R.

INTRODUCTION

Klebsiella pneumoniae (K. pneumoniae) is one of the main causes of either hospital or community-acquired infections. It causes a wide variety of infections including; infections of respiratory tract, urinary tract, blood stream and intra-abdominal abscesses (1).

For a long time, carbapenems have been the last antibiotic choice for the management of patients infected with extended spectrum beta lactamase (ESBL)-producing K. pneumoniae (2). Unfortunately, the widespread use of antibiotics, its self-administration by patients and non-implementation of policies for proper and wise use of antibiotics in hospitals lead to emergence of carbapenem-resistant K. pneumoniae (CRKP) strains which are spreading worldwide (3, 4).

Non-susceptibility of K. pneumoniae to carbapenems is mediated by a variety of mechanisms including production of one or more of carbapenemase enzymes, impaired permeability of outer membrane together with hyperproduction of β-lactamases (ESBL or AmpC), and antibiotic efflux across the outer membrane (5).

The rapid rise of carbapenem resistance among K. pneumoniae together with the availability of only few choices of antimicrobial therapy for treating infections caused by CRKP has led to the development of new antimicrobial agents to solve this problem. One of these antibiotics is imipenem/relebactam (IMP/R) (6).

Relebactam is a bicyclic diazabicyclooctane new β-lactamase inhibitor. Recently, in July 2019, IMP/R obtained approval by the Food and Drug Administration (FDA) for the treatment of complicated urinary tract and intra-abdominal infections (7). The ability of this recently developed antibiotic combination to restore activity of imipenem (IMP) could vary with the mechanism causing IMP non-susceptibility. Therefore, we aimed to elucidate to what extent the IMP/R combination can restore IMP susceptibility for a collection of IMP-NS K. pneumoniae isolates according to their mechanisms of IMP non-susceptibility. This study was approved from IRB committee in our faculty of medicine, Mansoura University.

MATERIALS AND METHODS

Bacterial isolates.

Clinical samples were collected from infected patients hospitalized in Mansoura University hospitals in the period from November 2017 to April 2019. Processing of all samples was done in microbiology laboratory of Specialized Medical Hospital at Mansoura University and in medical microbiology and immunology department at Mansoura faculty of medicine. Identification of organisms up to species level and preliminary determination of IMP minimum inhibitory concentration (MIC) were done by automated Vitek®2 compact system (bioMérieux, Marcy-l’Etoile, France).

The mechanism of IMP non-susceptibility among K. pneumoniae with IMP MIC ≥ 2 μg/ mL was determined firstly by genotypic detection of the commonest genes encoding carabapenemases, ESBLs and AmpC. The overproduction of ESBL and/or AmpC together with loss of porins could be responsible for carbapeneme non-susceptibility in isolates that harbor genes for ESBL and/or AmpC but are free of carbapenemase genes (8). Therefore, the mechanism of IMP non-susceptibility among carbapenemase-free isolates was investigated by analysis of gene expression of outer membrane proteins (Ompk35 & Ompk36), measurement of the efflux pump activity and phenotypic assays of the ESBL and/or AmpC overproduction.

Phenotypic assays of ESBL and AmpC enzymes.

Phenotypic detection of ESBL was done by Vitek2 system using AST-GN73 cards. AmpC overproduction was done by cefoxitin-cloxacillin combined disk test using discs of cefoxitin (30 μg) with and without cloxacillin (200 μg) as inhibitor of AmpC β-lactamase. AmpC production was considered if the inhibition zone of cefoxitin with cloxacillin disc was increased ≥ 4 mm than the inhibition zone of cefoxitin disk alone (9).

Molecular detection of β–lactamase genes.

DNA was extracted from all IMP-NS K. pneumoniae isolates by DNA extraction kits (QIAGEN, GmbH, Germany) according to manufacture instructions. Three sets of multiplex PCR assays were done to detect the most common genes of Carbapenemases including bla KPC , bla OXA48 and metallo-β-lactamases (bla NDM , bla IMP , bla VIM ), ESBL (bla CTXM , bla TEM , bla SHV ) and AmpC (MOX, CIT, DHA, ACC, EBC and FOX). Primers used for detection of all investigated genes are listed in Table 1. Amplification of the investigated genes was performed following protocols previously described by Poirel et al. for carbapenemase genes (10), by Lee et al. for ESBL genes (11) and by Pérez-Pérez and Hanson for AmpC genes (12).

Table 1.

Sequence of primers used in multiplex PCR

Target gene	Primer sequence (5′–3′)	Product size (bp)
16S rRNA	F:AATACCGCATAATGTCGC	622
	R:CCCTCGTTTGTCCATATCT
OmpK35	F:GCGTCCAATGTTGAAGGT	778
	R: TAATGTGGTTTCGCCGAC
OmpK36	F:CCGTAACTCTGATTTCTTCG	588
	R:TTAGTTGGACGACCTGCT
bla IMP	F:GGAATAGAGTGGCTTAAYTCTC	232
	R:GGTTTAAYAAAACAACCACC
bla VIM	F:GATGGTGTTTGGTCGCATA	390
	R:CGAATGCGCAGCACCAG
bla NDM	F:GGTTTGGCGATCTGGTTTTC	621
	R:CGGAATGGCTCATCACGATC
bla OXA-48	F:GCGTGGTTAAGGATGAACAC	438
	R:CATCAAGTTCAACCCAACCG
bla KPC	F:GTCTAGTTCTGCTGTCTTG	798
	R:CTTGTCATCCTTGTTAGGCG
bla TEM	F:GAGACAATAACCCTGGTAAAT	851
	R:AGAAGTAAGTTGGCAGCAGTG
bla CTX	F:GAAGGTCATCAAGAAGGTGCG	550
	R:GCATTGCCACGCTTTTCATAG
bla SHV	F:AAGATCCACTATCGCCAGCAG	231
	R:ATTCAGTTCCGTTTCCCAGCGG
bla MOX	F:GCT GCT CAA GGA GCA CAG GAT	520
	R:CAC ATT GAC ATA GGT GTG GTG C
bla CIT	F:TGG CCA GAA CTG ACA GGC AAA	462
	R:TTT CTC CTG AAC GTG GCT GGC
bla DHA	F:AAC TTT CAC AGG TGT GCT GGG	405
	R:CCG TAC GCA TAC TGG CTT TGC
bla ACC	F:AAC AGC CTC AGC AGC CGG TTA	346
	R:TTC GCC GCA ATC ATC CCT AGC
bla EBC	F:TCG GTA AAG CCG ATG TTG CGG	302
	R:CTT CCA CTG CGG CTG CCA GTT
bla FOX	F:AAC ATG GGG TAT CAG GGA GAT G	190
	R:CAA AGC GCG TAA CCG GAT TGG

F=Forward; R= Reverse; D=AorGorT; Y=CorT.

In brief, each multiplex PCR was performed in a total reaction volume of 25 μL that contains a hot start Taq DNA polymerase master mix (12.5 μL), DNA template adjusted to 50 ng/μl (2 μL), forward and reverse primer pairs specific for each analyzed gene (variable concentrations following the reference protocol) and nuclease free water (up to a total volume of 25 μL). Amplification was done in a thermal cylcler (Applied Biosystems) according to conditions previously described. The optimal annealing temperature was 52ºC for carbapenemase genes, 62ºC for ESBL genes and 64ºC for AmpC genes. Detection of the amplified DNA was done by electrophoresis using agarose gel stained with ethidium bromide at 100V for 1 h in 1× TAE buffer. The separated bands were visualized by UV transillmuinator using 100 bp and 50 bp DNA ladder.

Determination of efflux pump activity.

Efflux activity was detected by MIC reduction assay using efflux pump inhibitor; carbonyl-cyanide3-chlorophenyl-hydrazine (CCCP) (HiMedia). IMP MIC of the tested isolate was determined by agar dilution method on Mueller-Hinton agar that contain 20 μg/mL of CCCP and compared to its MIC in the absence of CCCP. Significant pump activity with possible efflux pump gene overexpression is considered if there is a fourfold reduction in MIC of IMP in presence of CCCP (13).

Analysis of outer membrane proteins (Ompk35 & Ompk36) gene expression.

Gene expression analysis was performed by reverse transcription-PCR (RT-PCR) using OneStep RT-PCR Kit (QIAGEN). Extraction of RNA was done using an RNeasy Protect Bacteria MiniKit (QIAGEN) according to the manufacturer’s instructions. Normalization of mRNA expression levels in different strains was performed by 16S rRNA. Reverse transcription was done at 50ºC for 30 min, then initial PCR activation achieved at 95ºC for 15 min. DNA amplification cycles included; denaturation at 94ºC (1 min), then annealing at 48ºC (1 min) followed by extension at 72ºC (1 min). Cycles was repeated 40 times, thereafter, the final extension lasted for 10min at 72ºC (14). The carbapenem-susceptible K. pneumoniae ATCC 13883 (wild-type OmpK35 and OmpK36) was used as a control strain.

IMP and IMP/R susceptibility testing.

Susceptibility of all K. pneumoniae isolates to both IMP and IMP/R with determination of MIC was performed by broth microdilution (BMD) method simultaneously on the same day according to CLSI recommendations (15). P. aeruginosa ATCC 27853 and Escherichia coli ATCC 25922 strains were used as for quality control. Imipenem and relebactam powder were purchased from Sigma-Aldrich (St. Louis, USA). Doubling dilutions of IMP starting from 0.03 to 128 μg/mL with a relebactam fixed at concentration of 4 μg/Ml were used. MIC interpretation of both IMP and IMP/R was done according to CLSI guidelines (15). For IMP, K. pneumoniae was considered susceptible, intermediate and resistant if with MIC ≤1 μg/Ml, equal 2 μg/ mL and ≥4 μg/mL, respectively. For IMP/R, K. pneumoniae was considered susceptible, intermediate and resistant if with MIC ≤1/4, equal 2/4 and ≥4/4, respectively.

Data analysis.

Statistical analysis was done using Statistical Package for Social Sciences (SPSS, version 22; Chicago, IL, USA). Categorical variables were described in numbers and percentages.

RESULTS

One hundred and forty of IMP-NS K. pneumoniae isolates were collected from bacterial cultures. These isolates were obtained from cultures of urine (n=67), blood (n=54) and ascetic fluid (n=19).

Beta-lactamase profile of IMP/NS K. pneumoniae.

Multiplex PCR analysis categorized the studied IMP-NS isolates into 120 (85.7%) carbapenemase producers and 20 (14.3%) non-carbapenemase producers. Carbapenemases were detected either alone in 102 (72.8%) isolates or in combination with other β-lact- amases in 18 (12.8%) isolates; 12 (8.5%) isolates coproduced also ESBLs and 6 (4.3%) isolates harbored also both ESBLs and AmpC. The most common detected carbapenemase gene was bla KPC (n=58,41.4%); detected either alone in 44 (31.4%) isolates or together with other carbenemases in 14 (10%) isolates. MBL genes (bla NDM , bla IMP , bla VIM ) were detected in 48 (34.3%) isolates; alone in 36 (25.7%) isolates and in combination with other carbapenemases in 12 (8.6%) isolates. The least prevalent carbapenemase gene was bla OXA-48 ; found alone in 8 (5.7%) isolates and coproduced with other cabapenemase in 6 (4.3%) isolates. The twenty carbapenemase non-producer isolates were categorized as 10 (7.1%) ESBL-only producers, 6 (4.3%) AmpC-only producers and 4 (2.9%) co-producers of both ESBL and AmpC Table 2.

Table 2.

MIC distribution of imipenem and imipenem with relebactam in relation to β- lactamase type in K. pneumoniae isolates

B- lactamase type (n)	AB	No of isolates (cumulative %) inhibited at MIC (μg/mL)													MIC50	MIC90	S%
		0.03	0.06	0.12	0.25	0.5	1	2	4	8	16	32	64	128
Carbapenemase positive (120)	I	0	0	0	0	0	0	10 (8.3)	16 (21.7)	22 (40)	24 (60)	22 (78.3)	16 (91.7)	10 (100)	16	64	0
	I/R	16 (13.3)	12 (23.3)	14 (35)	0	4 (38.3)	4 (41.7)	6 (46.7)	10 (55)	14 (66.7)	16 (80)	14 (91.7)	10 (100)	0	4	32	41.7
Carbapenemase-only producers (102)	I	0	0	0	0	0	0	10 (9.8)	16 (25.5)	20 (45.1)	20 (64.7)	16 (80.4)	14 (94.1)	6 (100)	16	64	0
	I/R	16 (15.7)	12 (27.5)	14 (41.2)	0	4 (45.1)	4 (49)	6 (54.9)	10 (64.7)	10 (74.5)	12 (86.3)	8 (94.1)	6 (100)	0	2	32	49
bla KPC only producers (44)	I	0	0	0	0	0	0	8 (18.2)	8 (36.4)	10 (59.1)	8 (77.3)	6 (90.9)	2 (95.5)	2 (100)	8	32	0
	I/R	16 (36.4)	10 (59.1)	12 (86.4)	0	2 (90.9)	4 (100)	0	0	0	0	0	0	0	0.06	0.5	100
MBL only producers (36)	I	0	0	0	0	0	0	0	8 (22.2)	8 (44.4)	10 (72.2)	4 (83.3)	6 (100)	0	16	64	0
	I/R	0	0	0	0	0	0	0	8 (22.2)	8 (44.4)	10 (72.2)	4 (83.3)	6 (100)	0	16	64	0
bla OXA-48 producers (8)	I	0	0	0	0	0	0	2 (25)	0	2 (50)	0	2 (75)	0	2 (100)	8	128	0
	I/R	0	2 (25)	2 (50)	0	0	0	0	0	0	2 (75)	2 (100)	0	0	0.12	32	50
bla KPC& bla OXA-48 coproducers (2)	I	0	0	0	0	0	0	0	0	0	2 (100)	0	0	0	16	16	0
	I/R	0	0	0	0	2 (100)	0	0	0	0	0	0	0	0	0.5	0.5	100
bla KPC& MBL coproducers (8)	I	0	0	0	0	0	0	0	0	0	0	4 (50)	4 (100)	0	32	64	0
	I/R	0	0	0	0	0	0	2 (25)	0	2 (50)	0	4 (100)	0	0	8	32	0
bla KPC& MBL & bla OXA-48 coproducers (4)	I	0	0	0	0	0	0	0	0	0	0	0	2 (50)	2 (100)	64	128	0
	I/R	0	0	0	0	0	0	2 (50)	0	0	2 (100)	0	0	0	2	16	0
ESBL & Carbapenemase coproducers (12)	I	0	0	0	0	0	0	0	0	2 (16.7)	4 (50)	6 (100)	0	0	16	32	0
	I/R	0	0	0	0	0	0	0	0	4 (33.3)	4 (66.7)	4 (100)	0	0	16	32	0
ESBL & AmpC & Carbapenemase coproducers (6)	I	0	0	0	0	0	0	0	0	0	0	0	2 (33.3)	4 (100)	128	128	0
	I/R	0	0	0	0	0	0	0	0	0	0	2 (33.3)	4 (100)	0	64	64	0
Carbapenemase negative (20)	I	0	0	0	0	0	0	4 (20)	0	4 (40)	4 (60)	0	6 (90)	2 (100)	16	64	0
	I/R	2 (10)	2 (20)	0	2 (30)	2 (40)	6 (70)	4 (90)	2 (100)	0	0	0	0	0	1	2	60
ESBL only producers (10)	I	0	0	0	0	0	0	2 (2)	0	2 (40)	2 (60)	0	4 (100)	0	16	64	0
	I /R	0	0	0	0	2 (20)	4 (60)	2 (80)	2 (100)	0	0	0	0	0	1	4	60
AmpC only producers (6)	I	0	0	0	0	0		2 (33.3)		2 (66.7)	2 (100)	0	0	0	8	16	0
	I /R	2 (33.3)	2 (66.7)	0	2 (100)	0	0	0	0	0	0	0	0	0	0.06	0.25	100
ESBL & AmpC coproducers (4)	I	0	0	0	0	0	0	0	0	0	0	0	2 (50)	2 (100)	64	128	0
	I /R	0	0	0	0	0	2 (50)	2 (100)	0	0	0	0			1	2	50
Total isolates (140)	I	0	0	0	0	0	0	14 (10)	16 (21.4)	26 (40.0)	28 (60.0)	22 (75.7)	22 (91.4)	12 (100)	16	64	0
	I /R	18 (12.9)	14 (22.9)	14 (32.9)	2 (34.3)	6 (38.6)	10 (45.7)	10 (52.9)	12 (61.4)	14 (71.4)	6 (82.9)	14 (92.9)	10 (100)	0	2	32	45.7

AB= Antibiotic; I= Imipenem alone; I/R= Imipenem with relebactam; S=Susceptible; MIC= Minimal inhibitory concentration; MBL= Metallo-β-latamase, ESBL= Extended-spectrum-β-lactamase.

Porins expression and efflux pump activity in carbapenemase-negative IMP-NS isolates.

Expression loss of OmpK35 and/or OmpK36 in combination with production of ESBL and/or AmpC was detected in 18/20 (90%) of carbapenemase-negative isolates. Out of ESBL-only producers, 8 (80%) showed expression loss of OmpK35 and/or OmpK36 without efflux pump hyperactivity while the remaining two isolates showed only efflux pump hyperactivity without any expression loss of outer membrane proteins. All six AmpC-only producers showed expression loss of both OmpK35 and OmpK36 and none of them showed efflux pump hyperactivity. All four ESBL and AmpC coproducers showed expression loss of OmpK35 and/or OmpK36 without efflux pump hyperactivity with no demonstrable efflux pump hyperactivity, Table 3.

Table 3.

Effect of relebactam on susceptibility to IMP in carbapenemase-negative IMP-resistnt K. pneumoniae isolates in relation to ESBL & AmpC overproduction, expression of porin genes and efflux pump activity.

Isolate No	Phenotypic detection of		Gene expression of		Efflux activity	MIC (μg/mL)
	ESBL	AmpC	OmpK35	OmpK36		IMP	IMP/R
7	+	−	Lost	Present	No	2	0.5
16	−	+	Lost	Lost	No	2	0.03
24	+	+	Lost	Present	No	64	1
35	+	−	Lost	Lost	No	16	1
40	−	+	Lost	Lost	No	2	0.03
42	+	−	Present	Present	Yes	64	4
50	−	+	Lost	Lost	No	16	0.25
55	+	−	Lost	Present	No	2	0.5
60	+	−	Lost	Lost	No	64	2
64	+	−	Lost	Present	No	8	1
69	+	+	Lost	Lost	No	128	2
72	−	+	Lost	Lost	No	8	0.06
80	+	−	Lost	Present	No	8	1
89	−	+	Lost	Lost	No	16	0.25
102	+	−	Lost	Lost	No	64	2
115	−	+	Lost	Lost	No	8	0.06
120	+	+	Lost	Present	No	64	1
129	+	−	Lost	Lost	No	16	1
132	+	+	Lost	Lost	No	128	2
138	+	−	Present	Present	Yes	64	4

IMP= Imipenem alone; IMP/R= Imipenem with relebactam; MIC= Minimal inhibitory concentration, ESBL= Extended-spectrum-β-lactamase

In vitro activity of IMP and IMP/R by BMD method in relation to mechanism of IMP non-susceptibility.

Out of the collected 140 IMP/NS K. pneumoniae isolates; 14 (10%) showed intermediate susceptibility to IMP and 126 (90%) were IMP resistant with MIC ranged from 4 to 128 μg/mL. Addition of relebactam restored susceptibility to IMP in 64 (45.7%) of all IMP-NS K. pneumoniae isolates with reduction of MIC 50 8 folds from 16 to 2 μg/mL and lowering of MIC 90 2 folds from 64 to 32 μg/mL.

In accordance to β-lactamase type, relebactam restored IMP activity in 50/102 (49%) of carbapenemase-only producing isolates, in 6/6 (100%) of AmpC-only producing isolates, in 6/10 (60%) of ESBL-only producing isolates and in 50% of isolates co-producing both ESBL & AmpC. However, none of isolates co-producing carbapenemase in combination with either ESBL alone or with both ESBL and AmpC rendered susceptible to IMP by relebactam, Table 2.

Regarding the effect of carbapenemase type on the activity of relebactam, addition of relebactam restored IMP susceptibility in 100% of KPC-producers, in 50% of OXA-48- producers and in 100% of isolates harboring both KPC and OXA-48 enzymes. However, none of isolates harboring MBL enzymes either alone or in combination with either bla KPC or bla OXA-48 rendered susceptible to IMP by addition of relebactam. Reduction of MIC 90 was 64 folds in bla KPC producers and 32 folds in isolates producing bla OXA-48 ; either alone or in combination with bla KPC (Table 2).

For the in vitro effect of IMP/R on IMP-NS carbapenemase-negative K. pneumoniae, none of the two isolates with only efflux pump hyperactivity rendered IMP susceptible by addition of relebactam. Out of 18 isolates with expression loss of OmpK35 and/ or OmpK36 in combination with production of ESBL and/or AmpC, IMP susceptibility was restored in 14 (77.7%) of them. These IMP/R susceptible isolates included; all AmpC-only producers, 6/8 (75%) of ESBL-only producers, 2 (50%) of ESBL and AmpC coproducers, (Table 3).

DISCUSSION

Many studies investigated the in-vitro activity of IMP/R against carbapenemase mediated IMP non-susceptibility in Gram-negative bacilli. However, only limited studies investigated the in-vitro activity of this new antibiotic against different mechanisms of carbapenem non-susceptibility in K. pneumoniae. We compared the in-vitro effect of IMP/R versus IMP alone in a collection of IMP-NS K. pneumoniae isolates harboring different mechanisms of IMP non-susceptibility.

Multiplex PCR performed for detection of the type of β-lactamases responsible for the reduced IMP susceptibility among the collected isolates revealed that IMP non-susceptibility are mediated by carbapene-amses in 120 (85.7%) of isolates. Whereas the IMP non-susceptibility in the remaining 20 (14.3%) isolates are mediated by mechanisms other than carbapenemases. This is similar to Gomez-Simmonds et al. who reported that 88.9% and 10.4% of their carbapenam resistant Enterobactericeae are carbapenemase positive and negative, respectively (16).

To date of performing this analysis, susceptibility breakpoints of IMP/R were not established by CLSI 2019. Therefore, breakpoints of IMP published in CLSI 2019 were used for interpretation of both IMP and IMP/R susceptibility results. However, CLSI published in March 2021 the interpretive criteria of IMP/R by BMD method (17). Fortunately, these recently published criteria are the same as we used in the interpretation, so it did not affected on our findings.

In the present study, the most potent activity of IMP/R was against bla KPC -carrying isolates. Relebactam restored susceptibility to IMP in 100% of the isolates that harbor carbapenemase of bla KPC -type. This agrees with other studies reporting that relebactam restored imipenem susceptibility in all K. pneumoniae with carbapenemases of bla KPC-2 and bla KPC-3 typee (18).

On the other hand, addition of relebactam did not restore IMP susceptibility in any of MBL-only producing K. pneumoniae isolates. This is consistent with previous study that evaluated IMP/R susceptibility among 113 isolates of K. pneumoniae harboring class B-metallo-β-lactamases, and found that none of them rendered IMP susceptible by relebactam regardless of MβL type (6).

Regarding the effect of IMP/R on the bla OXA-48 -producing K. pneumoniae isolates, half of the evaluated IMP-NS bla OXA-48 producers were found to be susceptible to IMP/R. This is in accordance with Schmidt-Malan et al. who found that 50% of their bla OXA-48 -positive isolates had IMP/R MICs of 1 μg/mL (19). However, other studies reported that IMP/R has no activities against isolates with class D-carbapenemase (7, 18). Therefore, it seems that relebactam does not consistently inhibit IMP in isolates producing bla OXA-48 -carbapenemases.

In our study, investigation of porins gene expression, efflux pump as well as ESBL and AmpC overproduction in carbapenemase-negative IMP non-susceptible isolates showed that IMP non-susceptibility in 90% of these isolates could be caused by combined loss of either one or both porins with overproduction of ESBL/or AmpC. This is in line with Hamzaoui et al. who proved that loss of porins in K. pneumoniae can extend resistance spectrum mediated either by ESBLs or by plasmid-mediated AmpC enzymes to include carbapenems as well (8). This could be attributed to a residual capacity of ESBLs and AmpC enzymes to hydrolyse carbapenems, albeit at a very low efficiency (20).

Regarding the impact of relebactam on non-susceptibility to IMP caused by overproduction of ESBL and/or AmpC with loss or alteration of outer membrane proteins, we found that all AmpC-only producers, 75% of ESBL-only producers and 50% of ESBL and AmpC coproducers rendered susceptible to IMP by addition of relebactam. Similarly, Haidar et al. demonstrated a modest potentiation of IMP activity against carbapenem-resistant Enterobacteriaceae isolates producing ESBL and/or AmpC enzymes with membrane impermeability by addition of relebactam (21). Also, Gomez-Simmonds et al. found that 88% of their carbapenemase-negative Enterobactericeae carrying either bla AmpC or bla CTX-M-15 genes rendered IMP susceptible by relebactam (16). Moreover, relebactam has been found previously to restore the in-vitro activity of imipenem activity against P. aeruginosa that are carbapenem-resistant due to impermeability arising from porins loss combined with AmpC expression (6).

However, we observed that relebactam could not decrease IMP MIC to a susceptible level in six carbapenemase-negative isolates. Two of these isolates were ESBL-only producers expressing both porins but with efflux pump activity. Since relebactam is not subject to efflux (22), therefore, IMP/R resistance in these two isolates could be due to functional mutations in major OmpK36 porin (21). The other four isolates that showed non-susceptibility to IMP/R (MIC=2) were two ESBL-only producers and two coproducers of AmpC and ESBL showing expression loss of both porins. Therefore, it is possible that these isolates is expressing ESBL or AmpC in high amount so that relebactam is unable to render these isolates susceptible to imipenem (22).

The overall rate of IMP/R activity against IMPNS K. pneumoniae in our study was low (45.7%) as compared to other studies that demonstrated higher activity of IMP/R against K. pneumoniae with rates of susceptibility exceeding 95% (16). This could be attributed to inclusion among the collected isolates in this study of considerably high percentage (34.3%) of K. pneumoniae isolates expressing β-lactamases, as MBLs, that not inhibited by relebactam.

CONCLUSION

Efficacy of relebactam in rendering IMP/NS K. pneumoniae susceptible to IMP varied according to the mechanism of IMP non-susceptibility. It had the highest activity on KPC carbapenemase and AmpC β-lactamse. However, it showed no activity on MBL genes (bla NDM , bla IMP , bla VIM ) and efflux pump over-activity. Rlebactam activity in IMP/NS K. pneumoniae with production of bla OXA-48 or with altered membrane permeability needs more study.