Activity of imipenem/relebactam against Pseudomonas aeruginosa producing ESBLs and carbapenemases

Abstract

Background

ESBL- and carbapenemase-producing Pseudomonas aeruginosa are prevalent in, for example, the Middle East, Eastern Europe and Latin America, though rarer elsewhere. Because P. aeruginosa readily mutate to become carbapenem resistant via loss of OprD, isolates producing ESBLs are often as broadly resistant as those producing carbapenemases. We hypothesized that: (i) relebactam might overcome class A carbapenemases directly in P. aeruginosa; and (ii) relebactam’s inhibition of AmpC, which gives a generalized potentiation of imipenem against the species, might restore imipenem susceptibility in OprD-deficient ESBL producers.

Methods

MICs were determined using CLSI agar dilution for P. aeruginosa isolates producing ESBLs, principally VEB types, and for those producing GES-5, KPC and other carbapenemases.

Results

Relebactam potentiated imipenem by around 4–8-fold for most P. aeruginosa isolates producing VEB and other ESBLs; however, MICs were typically only reduced to 4–16 mg/L, thus mostly remaining above EUCAST’s susceptible range and only partly overlapping CLSI’s intermediate range. Strong (approx. 64-fold) potentiation was seen for isolates producing KPC carbapenemases, but only 2-fold synergy for those with GES-5. Predictably, potentiation was not seen for isolates with class B or D carbapenemase activity.

Conclusions

Relebactam did potentiate imipenem against ESBL-producing P. aeruginosa, which are mostly imipenem resistant via OprD loss, but this potentiation was generally insufficient to reduce imipenem MICs to the clinical range. Imipenem resistance owing to KPC carbapenemases was reversed by relebactam in P. aeruginosa, just as for Enterobacterales.

Introduction

Most resistance to β-lactams in Pseudomonas aeruginosa in Western Europe and North America arises via mutations that up-regulate efflux mediated by MexAB-OprM and other pumps, derepress AmpC-β-lactamase expression, or cause inactivation of the ‘carbapenem-specific’ porin OprD.¹ Only small minorities of isolates owe resistance to acquired ESBLs or carbapenemases, though isolates producing these enzymes are prevalent in Eastern Europe,² Russia,³ the Middle East^4,5 and Latin America.^6,7 Unlike for Enterobacterales, where CTX-M, SHV and TEM types predominate, most ESBLs in P. aeruginosa are VEB and PER types.^5,8 Among carbapenemases, MBLs (class B) dominate in most countries,⁹ but KPC enzymes (class A) are prevalent in Colombia and the Caribbean,^6,10 with reports of OXA-48-like (class D) enzymes from India and Turkey.^11,12 GES enzymes (class A) occur too; some are ESBLs, but others are carbapenemases.¹³ Because P. aeruginosa readily mutate to become carbapenem resistant (via inactivation of OprD porin),¹ ESBL producers are frequently as broadly resistant as carbapenemase producers. This is in contrast to ESBL-producing Enterobacterales, which mostly remain susceptible to carbapenems.

New antibiotics offer the potential to overcome some of these challenges. Ceftolozane/tazobactam remains active against most P. aeruginosa with mutational resistance to β-lactams, regardless of mechanism, and ceftazidime/avibactam is active against those with derepressed AmpC or inactivated OprD. However, both of these cephalosporin/inhibitor combinations lack reliable activity against strains producing ESBLs or carbapenemases.^14,15 We hypothesized that imipenem/relebactam might have greater potential in these latter cases. First, and most simply, we posited that relebactam should inhibit class A carbapenemases, restoring the activity of imipenem. Secondly, we reasoned that there was a potential for activity against carbapenem-resistant ESBL producers, as these owe their carbapenem resistance to a combination of loss of OprD loss and AmpC activity, meaning that, if relebactam inhibited the AmpC enzyme, the imipenem MIC might be lowered into the clinical range. In context, it is vital to understand that the chromosomal AmpC β-lactamase of P. aeruginosa, ubiquitous in the species, ordinarily provides a small degree of protection against imipenem, such that imipenem/relebactam MICs for WT P. aeruginosa are approx. 4-fold lower than those of imipenem alone. This differential extends to 8-fold for isolates lacking OprD.^16,17 Potentiation of imipenem was likewise seen with other AmpC inhibitors that were not developed^18,19 and mutational loss of AmpC restores imipenem susceptibility in OprD-deficient strains.¹⁶ Thus, somewhat counterintuitively, an AmpC inhibitor may overcome what is considered to be an ‘impermeability-mediated’ resistance.

Panels of P. aeruginosa isolates producing ESBLs and carbapenemases were used to test these hypotheses; these were selected from submissions to PHE and reflected the distribution of ESBLs and carbapenemases seen over the past decade, except that we deliberately under-represented MBL producers, since there was no reasonable expectation that imipenem/relebactam would be active against them.

Materials and methods

Isolates

Isolates were non-replicate submissions of P. aeruginosa producing ESBLs or non-metallo carbapenemases referred to the PHE Antimicrobial Resistance and Healthcare Associated Infections (AMRHAI) Reference Unit between 2012 and 2019. β-Lactamase genes were identified by PCR (for primers used, which were refined over time, see Table S1, available as Supplementary data at JAC Online) or by WGS using Illumina methodology, which was undertaken when an isolate had a phenotype suggesting an ESBL or carbapenemase, but routine PCR failed to identify a corresponding gene. When WGS was performed, reads from each genome were assembled de novo and screened for antimicrobial resistance genes using Blast software and PHE’s in-house Genefinder bioinformatics pipeline.¹⁴ ESBL producers variously expressed VEB (n = 97), PER (n = 9), GES ESBL (n = 7, one each with GES-1 and GES-7, three with GES-9 and two with GES-26), SHV (n = 2, one each with SHV-5 and SHV-12) and CTX-M-15 (n = 1) ESBLs. Carbapenemase producers variously expressed GES-5 (n = 37), OXA-48-like (n = 4, one with known OXA-181), MBL (n = 11, five with NDM enzymes, five with VIM types and one with both) and KPC (n = 2, unsequenced) enzymes. Variable-number tandem repeat (‘VNTR’) typing or WGS data were available for most isolates producing VEB and GES enzymes, with STs thereby deduced.²⁰ Among the 97 isolates producing VEB ESBLs, 75 (from at least 27 different hospitals) belonged to ST357 or its single-locus variants (SLVs) and 8 belonged to ST654 or its SLVs; the remainder were sporadic types (n = 6) or not typed (n = 8). Among the 37 isolates producing GES-5 enzymes, 25 (from seven hospitals) belonged to ST235 and 2 belonged to the ‘Nottingham strain’;²¹ 10 were not typed.

MIC determinations

MICs were determined using CLSI agar dilution,²² with all β-lactamase inhibitors used at 4 mg/L. Imipenem, relebactam and ceftolozane were from Merck, Sharp and Dohme (Hoddesdon, UK). Imipenem, meropenem, ceftazidime, tobramycin, amikacin, gentamicin, colistin and tazobactam were from Merck KGaA (Gillingham, UK) and avibactam was from Pfizer.

Results

MIC distributions of imipenem/relebactam and its comparators are shown in Table 1; fold reductions in imipenem MIC achieved with relebactam are shown in Table 2. Susceptibility data for β-lactams are reviewed against current EUCAST and CLSI breakpoints in Table 3.

Table 1.

MIC distributions for P. aeruginosa isolates producing ESBLs and carbapenemases

	Number of isolates with indicated MIC (mg/L)
	≤0.25	0.5	1	2	4	8	16	32	64	128	>128
Imipenem
ESBLs
VEB (97)				3	3	3	9	32	47
PER (9)				2		2	2	3
GES ESBL (7)				1	2	1	3
SHV (2)			1					1
CTX-M (1)					1
carbapenemases
GES-5 (37)									15	22
KPC (2)										2
MBL (11)									1	1	9
OXA-48-like (4)									2	1	1
susceptible (4)		3	1
Meropenem
ESBLs
VEB (97)		4	2			2	4	47	24	14
PER (9)						2	1	3		3
GES ESBL (7)					1	3	3
SHV (2)						1		1
CTX-M (1)						1
carbapenemases
GES-5 (37)										11	26
KPC (2)									1	1
MBL (11)									1	2	8
OXA-48-like (4)									2	1	1
susceptible (4)	3			1
Imipenem/relebactam
ESBLs
VEB (97)		5	3	2	16	51	20
PER (9)		1		1	3	3	1
GES ESBL (7)		1	1	2	2	1
SHV (2)		1				1
CTX-M (1)			1
carbapenemases
GES-5 (37)								7	29	1
KPC (2)			1	1
MBL (11)									1	1	9
OXA-48-like (4)									2	2
susceptible (4)	3	1
Ceftazidime
ESBLs
VEB (97)									1		96
PER (9)											9
GES ESBL (7)									2		5
SHV (2)											2
CTX-M (1)										1
carbapenemases
GES-5 (37)						3	14	17	2	1	1
KPC (2)											2
MBL (11)									1	3	7
OXA-48-like (4)				3							1
susceptible (4)			3	1
Ceftazidime/clavulanate
ESBLs
VEB (97)					5	20	36	21	8	3	4
PER (9)						1	2	1	2		3
GES ESBL (7)								1	1	1	4
SHV (2)								1		1
CTX-M (1)								1
carbapenemases
GES-5 (37)						3	23	8	3
KPC (2)										1	1
MBL (11)									3	2	6
OXA-48-like (4)				2	1						1
susceptible (4)			2	2
Ceftazidime/avibactam
ESBLs
VEB (97)					1	1	4	18	44	20	9
PER (9)					1		1	3	1		3
GES ESBL (7)					2	1		3		1
SHV (2)							1		1
CTX-M (1)								1
carbapenemases
GES-5 (37)				3	26	6	2
KPC (2)						1	1
MBL (11)									3	2	6
OXA-48-like (4)				3				1
susceptible (4)			3	1
Ceftolozane/tazobactam
ESBLs
VEB (97)					1		2	1	1	8	84
PER (9)								1		2	5
GES ESBL (7)							2		2	1	2
SHV (2)								1	1
CTX-M (1)								1
carbapenemases
GES-5 (37)					3	17	15	2
KPC (2)										2
MBL (11)											11
OXA-48-like (4)			3								1
susceptible (4)	1	3
Piperacillin/tazobactam
ESBLs
VEB (97)							7	15	35	20	20
PER (9)								2	2	1	4
GES ESBL (7)							2	1	3		1
SHV (2)											2
CTX-M (1)											1
carbapenemases
GES-5 (37)								3	26	7	1
KPC (2)											2
MBL (11)									1	1	9
OXA-48-like (4)								2	1	1
susceptible (4)		1	1		2
Tobramycin
ESBLs
VEB (97)			1	2	1	5	10		12	32	34
PER (9)			2	1		1		1			4
GES ESBL (7)									1		6
SHV (2)								2
CTX-M (1)											1
carbapenemases
GES-5 (37)						1	17	6	1		12
KPC (2)			1							1
MBL (11)								2			9
OXA-48-like (4)										3	1
susceptible (4)	1	1		2
Gentamicin
ESBLs
VEB (97)				3	7	6	9	19	4	3	46
PER (9)					1	2			2		4
GES ESBL (7)									1		6
SHV (2)									1	1
CTX-M (1)											1
carbapenemases
GES-5 (37)					1		17	4		1	14
KPC (2)					1						1
MBL (11)					1					1	9
OXA-48-like (4)								2	1		1
susceptible (4)			1	1	1	1
Amikacin
ESBLs
VEB (97)						10	6	3	19	34	25
PER (9)					1	1	1	1		1	4
GES ESBL (7)							1		4	1	1
SHV (2)							1			1
CTX-M (1)										1
carbapenemases
GES-5 (37)			1		14	7	2		2	7	4
KPC (2)						1				1
MBL (11)								1	1		9
OXA-48-like (4)									1	2	1
susceptible (4)				1	1	2
Colistin
ESBLs
VEB (97)			2	87	4	1	3
PER (9)			1	7	1
GES ESBL (7)				7
SHV (2)		1		1
CTX-M (1)			1
carbapenemases
GES-5 (37)				34	2		1
KPC (2)				2
MBL (11)			2	9
OXA-48-like (4)				4
susceptible (4)				4

Table 2.

Potentiation of imipenem by relebactam for P. aeruginosa isolates with different modes of resistance

Mechanism	Number of cases with indicated fold reduction in MIC
	128	64	32	16	8	4	2	no change
ESBLs
VEB (97)				1	41	52	3
PER (9)					1	3	3	2
GES ESBL (7)						5	2
SHV (2)						1	1
CTX-M (1)						1
Carbapenemases
GES-5 (37)							28	9
KPC (2)	1	1
MBL (11)								11
OXA-48-like (4)						1		3
susceptible (4)							4

Table 3.

Proportions of isolates susceptible to newer β-lactamase inhibitor combinations and their parent compounds

	Definition, mg/L
	imipenem			imipenem/relebactam			ceftazidime		ceftazidime/avibactam				ceftolozane/tazobactam
	EUC, ≤4	CLSI-S, ≤2	CLDI-I, 4	EUC, ≤2	CLSI-Sa, ≤2	CLDI-Ia, 4	EUC, ≤8	CLSI-S, ≤8	CLDI-I, 16	EUC, ≤8	CLSI-S, ≤8	CLDI-I, NA	EUC, ≤4	CLSI-S, ≤4	CLDI-I, 8
ESBLs
VEB (97)	6.2%	3.1%	3.1%	10.3%	10.3%	16.5%	0%	0%	0%	2.1%	2.1%	NA	1.0%	1.0%	0%
PER (9)	2/9	2/9	0/9	2/9	2/9	3/9	0/9	0/9	0/9	1/9	1/9	NA	0/9	0/9	0/9
GES ESBL (7)	3/9	1/9	3/9	4/7	4/7	6/7	0/7	0/7	0/7	3/7	3/7	NA	0/7	0/7	0/7
SHV (2)	1/2	1/2	0/2	1/2	1/2	0/2	0/2	0/2	0/2	1/2	1/2	NA	0/2	0/2	0/2
CTX-M (1)	1/1	0/1	1/1	1/1	1/1	0/1	0/1	0/1	0/1	0/2	0/2	NA	0/1	0/1	0/1
Carbapenemases
GES-5 (37)	0%	0%	0%	0%	0%	0%	8.1%	8.1%	45.9%	94.6%	94.6%	NA	8.1%	8.1%	45.9%
KPC (2)	0/2	0/2	0/2	2/2	2/2	0/2	0/2	0/2	0/2	1/2	1/2	NA	0/2	0/2	0/2
MBL (11)	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	NA	0%	0%	0%
OXA-48-like (4)	0/4	0/4	0/4	0/4	0/4	0/4	3/4	3/4	0/4	3/4	3/4	NA	3/4	3/4	0/4
susceptible (4)	4/4	4/4	0/4	4/4	4/4	0/4	4/4	4/4	0.4	4/4	4/4	NA	4/4	4/4	0/4

EUC, EUCAST susceptible (S) and, if applicable, high-dose susceptible (I) pooled, on the basis that those agents with an I category should ordinarily be used at highest licensed doses against P. aeruginosa; CLSI-S and CLSI-I, CLSI susceptible and intermediate, respectively; NA, not applicable.

a

FDA breakpoint, now also adopted by CLSI, although publication of this remains pending.

The great majority of ESBL producers were resistant to imipenem and meropenem by all criteria, including >90% of those producing VEB enzymes and >66% of those producing other ESBLs. Since ESBLs do not attack carbapenems, such resistance must reflect other factors, putatively inactivation of OprD, as the near-universal mechanism of non-carbapenemase-mediated carbapenem resistance in P. aeruginosa.¹ All the carbapenemase-producing P. aeruginosa isolates were resistant to both carbapenems by all criteria.

Relebactam at 4 mg/L achieved 4–8-fold reductions in imipenem MICs for most isolates producing VEB ESBLs, and 2–4-fold reductions for those producing other ESBLs (Table 2). Nevertheless, MICs of the combination mostly remained around 4–16 mg/L, thus falling above EUCAST’s susceptible range (S ≤2 mg/L and R >2 mg/L) and with only a small overlap into the FDA’s intermediate range (S ≤2 mg/L and R >4 mg/L; values that it is understood also will be adopted by CLSI). MICs of the combination for P. aeruginosa isolates producing VEB enzymes belonging to the widespread ST357 lineage tended to exceed those for non-ST357 isolates (Figure 1), with five of the latter inhibited at <0.5 + 4 mg/L. Potentiation of imipenem against carbapenemase producers was generally ≤2-fold, including for isolates producing the GES-5 enzyme. Striking exceptions were the two isolates producing KPC enzymes, for which imipenem MICs were reduced from 128 to 1–2 mg/L. Unsurprisingly, there was no potentiation of imipenem for isolates producing class B (VIM or NDM) and class D (OXA-48-like) enzymes; these β-lactamases are not inhibited by relebactam.²³

Figure 1.

MICs of imipenem/relebactam for P. aeruginosa producing VEB ESBLs, according to whether these belonged or not to the widespread ST357 and its SLVs. Black, ST357 and SLVs (n = 75); grey, non-ST357 isolates (n = 14); 8 untyped isolates are excluded.

Ceftazidime was tested as a comparator, alone and combined with clavulanate and avibactam. Almost all ESBL producers were highly resistant to the unprotected cephalosporin, with MICs ≥128 mg/L, as were those producing MBLs or KPC enzymes. Isolates producing GES-5 enzymes were less resistant, with an MIC mode straddling 16–32 mg/L; three of four isolates producing the OXA-48-like enzyme were fully susceptible, with MICs of 2 mg/L. The remaining isolate with OXA-48-like activity was ceftazidime resistant and likely had a further mechanism. Clavulanate and avibactam reduced the MICs of ceftazidime for most ESBL-producing P. aeruginosa, though rarely sufficiently to bring MICs into clinical ranges. Thus, the modal MIC of ceftazidime for isolates producing VEB ESBLs fell from >128 mg/L with no inhibitor to 16 mg/L with clavulanate and to 64 mg/L with avibactam. Avibactam reduced the modal MIC for isolates producing GES-5 enzymes from 16–32 to 4 mg/L, with MICs for 35/37 isolates reduced to the EUCAST and CLSI breakpoint of ≤8 mg/L. Avibactam also reduced the MICs of ceftazidime for the two isolates with KPC carbapenemase activity from >128 to 8–16 mg/L.

The two commercial tazobactam combinations were also tested. In keeping with previous experience, ceftolozane/tazobactam was found to lack activity at accepted breakpoints against most ESBL and carbapenemase producers,¹⁴ exceptions being: (i) the 3 isolates producing OXA-48 carbapenemase also susceptible to ceftazidime; and (ii) 20/37 isolates producing GES-5 carbapenemase, which scored as susceptible or (mostly) intermediate by CLSI criteria, though only 3/37 were susceptible by EUCAST criteria. Piperacillin/tazobactam lacked activity against almost all the ESBL- and carbapenemase-producing P. aeruginosa isolates at 16 mg/L, corresponding to EUCAST’s I/R breakpoint and CLSI’s S/I breakpoint; it was active against 58.7% of VEB isolates at CLSI’s I/R breakpoint of 64 mg/L, though MICs of this level are associated with poor outcomes.²⁴

The final comparators were aminoglycosides and colistin. Resistance to tobramycin and gentamicin was seen for the great majority of isolates in all groups, whereas susceptibility to amikacin was frequent (64.9% by both CLSI and EUCAST criteria) only among those producing GES-5 carbapenemases. Colistin susceptibility appeared general in all groups, with only a few isolates found to be resistant; a caveat is that agar dilution was used for MIC determination, and this may occasionally miss resistance determined using broth microdilution.

Discussion

P. aeruginosa isolates producing ESBLs—principally VEB enzymes—and carbapenemases present major resistance challenges. Although uncommon in P. aeruginosa in Western Europe and North America, MBLs were present in 32% of carbapenem-resistant P. aeruginosa from Dubai²⁵ and 60% in Russia,³ where a successful ST235 lineage with VIM-2 enzyme has disseminated nationally. VEB ESBLs and various GES enzymes have repeatedly been shown to be widespread in P. aeruginosa in the Middle East,⁵ and in Mexico,²⁶ with dissemination of VEB types also reported in Bulgaria² and Thailand.²⁷

AMRHAI receives a steady flow of carbapenemase- and ESBL-producing P. aeruginosa, substantially from London private hospitals with international clienteles. Referrals with VEB enzymes were stable at 4–10 per annum up to 2016, then rose to 54 in 2017, 24 in 2018 and 66 in 2019, with the 2017 numbers augmented by an outbreak that saw eight representatives referred from one UK NHS site. Irrespective of where they are imported from, most belong to the ST357 lineage,²⁰ indicating an international clonal type, and this was strongly represented (75/97 isolates) in the present panel.

As illustrated here, and with previous collections,^14,20P. aeruginosa strains producing ESBLs typically are as broadly resistant as those producing carbapenemases, almost certainly owing to the ease with which carbapenem resistance develops via loss of OprD. Many ESBL- and carbapenemase-encoding plasmids simultaneously determine aminoglycoside-modifying enzymes or 16S rRNA methyltransferases, further expanding the spectrum of resistance.

We hypothesized that: (i) the general potentiation of imipenem by relebactam for P. aeruginosa, contingent on inhibition of AmpC, might overcome OprD-loss-mediated imipenem resistance in ESBL-producing isolates; and (ii) relebactam might directly overcome imipenem resistance mediated by class A carbapenemases.

Both hypotheses proved partially correct. For isolates producing VEB enzymes, relebactam achieved 4- or 8-fold MIC reductions for imipenem (Table 2), with 77/97 producers inhibited at 8 + 4 mg/L, corresponding to EUCAST’s high imipenem breakpoint from 2013–18. This positive finding is, however, negated by two developments since 2018, when this project was initiated. First, EUCAST’s imipenem breakpoint for P. aeruginosa was lowered from ≤4/>8 mg/L (pre-2018) to ≤2/>4 mg/L (2019) and then to ≤0.001/>4 mg/L (2020).²⁸ This latter change aimed to move the WT population of P. aeruginosa to ‘I’, underscoring EUCAST’s view that imipenem should ordinarily be used at a high dose (1 g q6h) for infections caused by P. aeruginosa. Secondly, imipenem/relebactam was licensed, by the EMA as well as the FDA, at 0.5 + 0.25 g q6h (i.e. half the licensed maximum dose for imipenem) and EUCAST consequently assigned a ≤2/>2 mg/L breakpoint. The FDA has proposed a ≤2/>4 mg/L breakpoint for imipenem/relebactam and this has now also been adopted by CLSI, which has an identical value for imipenem itself despite the higher maximum dosage.

Thus, although relebactam potentiated imipenem against isolates producing VEB enzymes and other ESBLs, MICs largely remained beyond the clinical range: only 10/97 (10.3%) isolates producing VEB enzymes were susceptible by EUCAST’s criteria, while 26/97 (26.8%) were susceptible or intermediate by FDA/CLSI criteria. Few isolates producing other ESBLs were included, these being extremely rare among AMRHAI submissions, but there was no suggestion of better performance than against those producing VEB types. Nor would differences based on ESBL type be expected, given that any general potentiation of imipenem against P. aeruginosa is contingent on inhibition of AmpC, not upon relebactam’s interactions with particular ESBLs.

Among carbapenemase producers, we predominantly tested isolates producing GES-5 enzymes, as the most prevalent class A carbapenemase in UK P. aeruginosa. MICs of imipenem alone were 64–128 mg/L and were only reduced by one doubling dilution by relebactam. Since GES-5 is a class A enzyme, this lack of potentiation is surprising and contrasts with the behaviour of avibactam, a structurally related diazabicyclooctane, which potentiated ceftazidime against these isolates. On the other hand, relebactam strongly potentiated imipenem against the two isolates producing KPC enzymes, with MICs reduced to the clinical range. While these carbapenemases are extremely rare among P. aeruginosa isolates in Europe and the UK (these two were the sole examples available to AMRHAI), they are prevalent in Colombia and several Caribbean islands.^6,7,10

Comparator results were in keeping with published data, except we found that ceftolozane/tazobactam was widely inactive against isolates with GES-5 activity (with MICs mostly 8–16 mg/L), whereas we previously found values of 2–4 mg/L, falling within EUCAST’s susceptible range.¹⁴ This difference may reflect the use of CLSI methodology with Mueller–Hinton agar, whereas BSAC methodology with Iso-Sensitest agar was used previously.

Given that imipenem/relebactam only narrowly failed to achieve activity against many isolates producing VEB and other ESBLs, with MICs that would have counted as intermediate under EUCAST’s 2018 imipenem breakpoints, it may be worth exploring whether pharmacodynamic exposure could usefully be increased with altered regimens. Although imipenem’s seizurogenic potential²⁹ is some constraint, the drug is licensed at regimens up to 1 g q6h when used alone (i.e. double the exposure of imipenem/relebactam), implying that some ‘headroom’ may exist. Likewise, although imipenem’s chemical instability complicates the use of extended infusions this should not be a barrier to increasing dosage frequency. In short, there may be routes to increase exposure and, given the paucity of alternatives against ESBL-producing P. aeruginosa, these deserve exploration.

Acknowledgements

We are grateful to Dr Katie Hopkins of PHE for helpful comments and discussion.

Funding

This work was funded by MSD as an Investigator-Initiated Project.

Transparency declarations

D.M.L.: Advisory Boards or ad-hoc consultancy (Accelerate, Allecra, Antabio, Centauri, Entasis, GlaxoSmithKline, Meiji, Melinta, Menarini, Mutabilis, Nordic, ParaPharm, Pfizer, QPEX, Roche, Shionogi, T.A.Z., Tetraphase, VenatoRx, Wockhardt and Zambon), paid lectures (Astellas, bioMérieux, Beckman Coulter, Cardiome, Cepheid, Merck/MSD, Menarini, Nordic, Pfizer and Shionogi) and relevant shareholdings or options (Dechra, GSK, Merck, Perkin Elmer, Pfizer and T.A.Z., amounting to <10% of portfolio value). All other authors: none to declare, but PHE’s AMRHAI Reference Unit has received financial support for conference attendance, lectures, research projects or contracted evaluations from numerous sources, including: Accelerate, Achaogen, Allecra, Amplex, AstraZeneca, AusDiagnostics, Basilea, Becton Dickinson, bioMérieux, Bio-Rad Laboratories, BSAC, Cepheid, Check-Points B.V., Cubist, Department of Health, Enigma Diagnostics, ECDC, Food Standards Agency, GenePOC™, GlaxoSmithKline, Helperby Therapeutics, Henry Stewart Talks, IHMA, Innovate UK, Kalidex Pharmaceuticals, Melinta Therapeutics, Merck Sharpe & Dohme, Meiji Seika, Mobidiag, Momentum Biosciences, Neem Biotech, NIHR, Nordic Pharma, Norgine Pharmaceuticals, Rempex Pharmaceuticals, Roche, Rokitan, Smith & Nephew, Shionogi, VenatoRx Pharmaceuticals, Wockhardt Ltd and the WHO.

Supplementary data

Table S1 is available as Supplementary data at JAC Online.