Assessment of Piperacillin-Tazobactam-Meropenem Synergy against Serine Carbapenemase-Producing Enterobacterales Using Time-Kill Assays

Synergy between piperacillin-tazobactam and meropenem against KPC-producing Klebsiella pneumoniae was recently demonstrated. We sought to test the combination against a broader range of serine carbapenemase producers.

ABSTRACT

Synergy between piperacillin-tazobactam and meropenem against KPC-producing Klebsiella pneumoniae was recently demonstrated. We sought to test the combination against a broader range of serine carbapenemase producers. We tested the combination against 10 KPC-producing Escherichia coli and 10 OXA-48 family-producing K. pneumoniae isolates. Antibiotic concentrations used are achievable in critically ill patients. The combination was synergistic against 7 of 10 KPC producers and 9 of 10 OXA-48 producers. There was no synergy detected in control isolates producing NDM-1.

INTRODUCTION

The global spread of multidrug-resistant bacteria is a significant threat to human health (1–3). Carbapenemase-producing Enterobacterales (CPE) are of particular concern because infections with these organisms are associated with considerable morbidity and mortality and can be challenging to treat because of limited therapeutic options (4–6). Based on the concept of double-carbapenem therapy (7), we recently investigated other beta-lactam combinations for in vitro synergy against KPC-producing Klebsiella pneumoniae and found that the combination of piperacillin-tazobactam and meropenem demonstrated an even greater degree of synergistic killing than ertapenem combined with meropenem (7). Furthermore, we demonstrated that all three beta-lactams were needed to achieve synergistic activity and that the effect could reduce CFU per milliliter by at least 4 log₁₀ in 58.3%. Motivated by these results, we sought to determine whether this effect was unique to KPC-producing K. pneumoniae or whether it could be found in other serine CPEs. To establish specificity of the synergy, we used NDM-1 producers as controls.

In isolates from patient specimens collected throughout Québec, Canada, we used time-kill assays to evaluate for synergy between piperacillin-tazobactam (piperacillin, Sigma-Aldrich, St. Louis, MO, USA; tazobactam, Cayman Chemicals, Ann Arbor, MI, USA) and meropenem (Fresenius Kabi, ON, Canada). Briefly, isolates were confirmed to be unique via pulsed-field gel electrophoresis (8) and to harbor the KPC, OXA-48-like, or NDM-1 genes by PCR at the Laboratoire de santé publique du Québec, our provincial reference laboratory. This laboratory tests for carbapenemase genes (bla_KPC, bla_NDM, bla_IMP, bla_VIM, bla_NMC, bla_SME, bla_{OXA-48 family}, and bla_GES) employing an in-house method that uses PCR techniques and primers based on previously published methods (9, 10). MICs were determined by broth microdilution per CLSI methodology (11), and the potencies of the antibiotics were confirmed serially over time using MIC determination of the quality control strain (12). The antibiotic concentrations used in the assays were selected in two ways. First, the concentrations were chosen to reflect those achievable in vivo in critically ill patients using optimal dosing strategies, i.e., 16/4 mg/liter for piperacillin-tazobactam and 8 mg/liter for meropenem (13, 14). Second, for organisms in which the meropenem MIC was <8 mg/liter, the time-kill assays were performed using concentrations of meropenem relative to the isolate’s MIC (1× MIC, 0.5× MIC, and 0.25× MIC). Each bacterial isolate was suspended in cation-adjusted Mueller-Hinton broth with piperacillin-tazobactam to achieve an initial bacterial concentration of 10⁶ CFU/ml. These suspensions were then incubated at 37°C and continuously mixed on a rotator. After 30 minutes of incubation, meropenem was added to the suspension in a minimal volume. The final concentration of meropenem was based on the meropenem MIC as described above. Subsequently, the suspension was reincubated at 37°C and continuously mixed, with colony counts performed per standard technique at 0.5, 4, 8, and 24 h. Synergy was defined as a ≥2-log₁₀-CFU/ml reduction at 24 h of exposure to the antibiotic combination compared with the single most active antibiotic, with the number of surviving organisms in the presence of the combination at >2 log₁₀ CFU/ml below the starting inoculum. Furthermore, it was required that piperacillin-tazobactam not appreciably affect the growth curve when used alone.

A total of 24 strains were included in this study, including 10 KPC-producing Escherichia coli isolates, 10 OXA-48 family-producing K. pneumoniae isolates, and 4 NDM-1-producing K. pneumoniae isolates. The MICs of the strains to piperacillin-tazobactam and meropenem and the results of the time-kill assays are listed in Table 1. Representative time-kill curves are shown in Fig. 1 for tested KPC producers and in Fig. 2 for tested OXA-48-like producers. The median MIC to meropenem was 3 mg/liter (interquartile range [IQR], 2 to 7 mg/liter) for the KPC-producing isolates and 0.5 mg/liter (IQR, 0.5 to 3.25 mg/liter) for the OXA-48-like-producing isolates. The MICs for the NDM-1 isolates to meropenem were either 16 or >32 mg/liter. For all isolates, the growth curves in the presence of piperacillin-tazobactam alone approximated those for the growth control. Synergy between piperacillin-tazobactam and meropenem was detected for 7 of 10 KPC-producing isolates and 9 of 10 OXA-48-like-producing isolates. No synergy was detected for the NDM-1-producing controls. For the 5 of 7 KPC-producing E. coli isolates whose MICs to meropenem were <8 mg/liter, the addition of piperacillin-tazobactam to meropenem at 0.25× the MIC demonstrated a ≥6-log₁₀-CFU/ml reduction at 24 h compared with meropenem monotherapy. The log₁₀-CFU/ml reduction for the combination relative to the single most active agent was less as meropenem monotherapy approached the MIC, indicating increasing activity of the monotherapy; however, synergy could still be demonstrated. A similar pattern was not observed with the OXA-48-like producers. Most of these isolates were susceptible to meropenem alone but also showed synergy between piperacillin-tazobactam and meropenem when the concentration of the meropenem was 0.5× or equal to the MIC. OXA-48-producing isolates did not demonstrate synergy between agents at 0.25× the meropenem MIC.

TABLE 1.

Reduction in log₁₀ CFU/ml of combinations relative to single most active agent

Isolate no.	Species	Carbapenemase type	MIC (mg/liter)b of:		Log10 reduction (CFU/ml)c with:
			MEM	PTZ	MEM (8 mg/liter)	1× MEM MIC	0.50× MEM MIC	0.25× MEM MIC
1	E. coli	KPC	4	>128/4		6	8	8
2a	E. coli	KPC	8	>128/4	3
3a	E. coli	KPC	16	>128/4	2
4	E. coli	KPC	2	128/4		5	5	6
5	E. coli	KPC	2	>128/4		4	8	10
6	E. coli	KPC	4	>128/4		4	5	7
7	E. coli	KPC	2	>128/4		2	2	2
8	E. coli	KPC	2	>128/4		5	6	8
9	E. coli	KPC	2	>128/4		7	5	1
10a	E. coli	KPC	8	>128/4	1
11	K. pneumoniae	OXA-48-like	0.5	>128/4		4	2	1
12	K. pneumoniae	OXA-48-like	0.5	>128/4		4	8	4
13a	K. pneumoniae	OXA-48-like	16	>128/4	2
14	K. pneumoniae	OXA-48-like	1	>128/4		5	5	-
15	K. pneumoniae	OXA-48-like	0.5	>128/4		7	10	2
16	K. pneumoniae	OXA-48-like	0.5	>128/4		8	1	1
17	K. pneumoniae	OXA-48-like	4	>128/4		5	3	6
18	K. pneumoniae	OXA-48-like	0.5	>128/4		7	2	1
19	K. pneumoniae	OXA-48-like	0.5	>128/4		5	6	1
20a	K. pneumoniae	OXA-48-like	16	>128/4	3
21a	K. pneumoniae	NDM-1	16	>128/4	0
22a	K. pneumoniae	NDM-1	>32	>128/4	0
23a	K. pneumoniae	NDM-1	16	>128/4	0
24a	K. pneumoniae	NDM-1	>32	>128/4	0

^aSynergy with meropenem 8 mg/liter was tested because the MICs for the isolates were at the limit of achievable in vivo meropenem concentrations in blood (see text for details).

^bMEM, meropenem; PTZ, piperacillin-tazobactam.

^cLog₁₀ reduction at 24 h relative to MEM monotherapy in presence of PTZ 16/4 mg/liter.

FIG 1.

Time-kill assay results for isolate 1. Meropenem (MER) concentrations were as follows: 1 mg/ml (0.25× MIC) (A), 2 mg/ml (0.5× MIC) (B), 4 mg/ml (1× MIC) (C). PTZ, piperacillin-tazobactam.

FIG 2.

Time-kill assay results for isolate 19. Meropenem (MER) concentrations were as follows: 0.125 mg/ml (0.25× MIC) (A), 0.25 mg/ml (0.5× MIC) (B), 0.5 mg/ml (1× MIC) (C). PTZ, piperacillin-tazobactam.

As antimicrobial resistance continues to evolve, it becomes critical to carefully evaluate existing drugs and determine how best to apply them, both alone and in combination. Although several carbapenemase inhibitors have come to market in the United States and Europe, there has already been a ceftazidime-avibactam shortage worldwide (15), and these drugs are not readily available or are prohibitively expensive in many countries. For instance, no carbapenemase inhibitor products are currently available in Canada. We previously demonstrated that the combination of piperacillin-tazobactam and meropenem showed synergistic killing against KPC-producing K. pneumoniae isolates. Here, we demonstrate that this effect is seen across a range of serine carbapenemase-producing organisms.

The biological mechanism for the increased activity of the drug combination against the serine carbapenemase producers remains unknown. Our results may suggest that the synergy observed involves inhibition of the carbapenemase enzymes themselves, given the specificity to the serine carbapenemases and the lack of efficacy against the metallo-beta-lactamase producers. Because the serine carbapenemase enzyme has a very wide range of beta-lactam substrates, it is plausible that the mechanism of synergy would likely involve saturation of the enzymatic active site such that the bactericidal beta-lactams can exert their effect. This is the proposed mechanism for double-carbapenem therapy (16). However, in our previous work, we showed that the synergy required the specific beta-lactams of meropenem, piperacillin, and tazobactam together; substituting for other carbapenems or cephalosporins resulted in reduction or removal of synergy (7). Whereas this could reflect specifics of the enzyme-substrate interactions and kinetics that favor the meropenem-piperacillin-tazobactam combination, this remains speculative because it has not been specifically investigated to date. We also acknowledge that there are numerous other commonalities between the two species tested, i.e., E. coli and K. pneumoniae. Consequently, there could be specific penicillin-binding protein inhibition or off-target effects that may also explain the results, although we would have expected similar synergy in the NDM-1 producers. One limitation of this study is that the control organisms, the NDM-1 producers, have much higher MICs to meropenem than the OXA-48-like and KPC producers. The higher potency of the enzymes for the substrates might be considered another limitation to synergy. Further work with a wider range of isolates will be required. We also acknowledge that using time-kill assays alone cannot characterize the causes of antibacterial synergy. The lack of universality of the synergy among the serine carbapenemase-producing isolates shown here and in our previous work suggests that some of the isolates possessed other resistance elements that could obviate the effectiveness of the combination. Further genetic characterization of these isolates is under way. Similarly, Enterobacter species are common producers of carbapenemases, particularly of KPC enzymes (17). However, members of this genus also frequently contain a number of Ambler class C enzymes, as well as various metallo-beta-lactamases. Because the theorized mechanism of synergy involves the suicide inhibition of the carbapenemase enzyme, the addition of other enzymes that may bind to meropenem, piperacillin, and tazobactam may preclude the bactericidal synergy observed here. Consequently, we did not test these species in this series of experiments because of the potential confounding of results.

In conclusion, our results further support the potential for combining piperacillin-tazobactam with meropenem in use against serine CPEs. Further in vivo studies and studies providing a mechanistic explanation for the combination are warranted.