Comparative Evaluation of the In Vitro Activities of WCK 5222 (Cefepime-Zidebactam) and Combination Antibiotic Therapies against Carbapenem-Resistant Pseudomonas aeruginosa

The in vitro activity of WCK 5222 (cefepime-zidebactam) was compared to that of several available combination therapies among 30 clinical carbapenem-resistant Pseudomonas aeruginosa (CRP) strains using gradient diffusion strips. The combinations included nonsusceptible β-lactams (cefepime, ceftolozane-tazobactam, and meropenem) with amikacin and fosfomycin.

ABSTRACT

The in vitro activity of WCK 5222 (cefepime-zidebactam) was compared to that of several available combination therapies among 30 clinical carbapenem-resistant Pseudomonas aeruginosa (CRP) strains using gradient diffusion strips. The combinations included nonsusceptible β-lactams (cefepime, ceftolozane-tazobactam, and meropenem) with amikacin and fosfomycin. WCK 5222 MICs ranged from 2 to 32 mg/liter, and 97% were ≤16 mg/liter, while 105/146 (72%) combinations demonstrated inhibition below established susceptibility breakpoints. WCK 5222 monotherapy may be preferred over the combinations assessed for CRP infections.

INTRODUCTION

Pseudomonas aeruginosa causes roughly 20% of hospital-acquired and ventilator-associated bacterial pneumonia cases in the United States, and 4 in every 10 such cases are fatal (1). Resistance to recently marketed β-lactam/β-lactamase inhibitor antibiotics such as ceftazidime-avibactam and ceftolozane-tazobactam (C/T) has been documented (2, 3), contributing to the growing concern surrounding these opportunistic pathogens. WCK 5222 (cefepime-zidebactam), a β-lactam/β-lactam enhancer antibiotic under development as a single intravenous formulation, has demonstrated in vitro (4) and in vivo (5) activity against multidrug-resistant (MDR) P. aeruginosa strains. Although zidebactam inhibits certain β-lactamases, the bactericidal effect of WCK 5222 is a result of complementary penicillin-binding protein inhibition (6).

We examined the in vitro activity of WCK 5222 against a selection of 30 clinical carbapenem-resistant (Carb^r) P. aeruginosa isolates (CRP) from U.S. hospitals and compared it to that of currently available antibiotic combinations. Since it is common for there to be a lag between market availability of a new antibiotic and its inclusion in automated antimicrobial susceptibility testing (AST) panels, diffusion methods conducted using a disk or gradient diffusion strip (GDS) often precede the reference method in the clinic. In addition, synergy tests using a checkerboard design or time-kill assays are not amenable to daily laboratory workflows (7). We therefore elected to exclusively use GDS (MIC test strip; Liofilchem, Roseto degli Abruzzi, Italy) in this assessment, and the comparator antibiotic combinations were carefully selected. Antipseudomonal β-lactams are routinely prescribed as a component of empirical therapy for nosocomial pneumonia in accordance with clinical guidance (8), but in cases where CRP is ultimately identified as the causative pathogen, they are unlikely to provide adequate coverage. To further increase clinical translation of this in vitro assessment, only combinations that included a nonsusceptible β-lactam were tested for synergistic interactions.

Bacterial inocula were prepared from pure, second subcultures of previously frozen stocks. Suspensions were uniformly applied to Mueller-Hinton II agar plates (BD BBL; Sparks Glencoe, MD). Synergy tests were performed by crossing a GDS containing cefepime (FEP), ceftolozane-tazobactam (C/T), or meropenem (MEM) at 90° angles with amikacin (AMK) and fosfomycin (FOF) at their respective MICs on inoculated agar plates (MTS synergy application system; Liofilchem). Each β-lactam was crossed with AMK and FOF if single GDS tests performed the day prior to synergy testing yielded the following MIC values: FEP, >8 mg/liter; C/T, >4/4 mg/liter; or MEM, >2 mg/liter (9). Fractional inhibitory concentration indices (FICIs) were calculated according to the equation:

$$\text{FICI}=\frac{{\text{MIC}}_{\beta \text{-lactam (with AMK or FOF)}}}{{\text{MIC}}_{\beta \text{-lactam alone}}}+\frac{{\text{MIC}}_{\text{AMK or FOF (with}\mathrm{\hspace{0.17em}\beta }\text{-lactam)}}}{{\text{MIC}}_{\text{AMK or FOF alone}}}$$

High off-scale MIC values were converted to the next 2-fold dilution and interactions were defined as: synergistic, FICI ≤ 0.5; additive, FICI > 0.5 to ≤ 1.0; indifferent, FICI > 1.0 to ≤ 4.0; and antagonistic, FICI > 4.0.

MICs of WCK 5222 were also determined using MIC test strips (Liofilchem). The product contains cefepime and zidebactam affixed in a 1:1 concentration gradient, which is consistent with the fixed 1:1 ratio established for broth microdilution tests (i.e., reference AST method) (9). Each inoculum, for single MIC tests and synergy evaluations, was prepared to a concentration of 1 × 10⁸ to 5 × 10⁸ CFU/ml and confirmed using serial dilution and plating techniques (Trypticase soy agar with 5% sheep blood; BD BBL, Franklin Lakes, NJ). To reduce the potential for overcalling synergy or underestimating WCK 5222 MICs, tests were repeated if the concentration was <1 × 10⁸ CFU/ml. P. aeruginosa ATCC 27853 served as the quality control (QC) organism for each antibiotic GDS; all QC tests were confirmed in range as specified by the manufacturer on each day of MIC and synergy testing.

MIC results are summarized in Table 1 and listed in Table S1 in the supplemental material. All MEM MICs were ≥8 mg/liter according to GDS, which was consistent with MEM resistance defined by previous susceptibility testing using the broth microdilution method (10). WCK 5222 demonstrated in vitro activity against the entire selection of CRP isolates (MIC range, 2 to 32 mg/liter), and zones of inhibited growth produced by WCK 5222 GDS were free from microcolonies or macrocolonies. It is noteworthy that the majority (23/30, 77%) of isolates displayed MIC values of ≤8 mg/liter (see Table S1 in the supplemental material), which demonstrates the ability of zidebactam to restore the susceptibility of cefepime. This activity is similar to that observed by Thomson et al. among MDR P. aeruginosa (n = 18) using agar dilution methods (cefepime-zidebactam [1:1] MIC₅₀, 8 mg/liter) (4). Recently, Monogue et al. assessed the in vivo efficacy of human-simulated exposures of WCK 5222 against cefepime- and meropenem-resistant P. aeruginosa strains (n = 16) in a neutropenic murine thigh infection model. Even for isolates with high WCK 5222 MICs of 16 to 32 mg/liter, which are consistent with the upper MIC range reported in the present study and by others (11), significant in vivo activity of WCK 5222 was observed (5).

TABLE 1.

Summary of gradient diffusion strip MICs for clinical carbapenem-resistant P. aeruginosa isolates (n = 30)

Antibiotic(s)	MIC (mg/liter)			Susceptible (%)
	Minimum	Maximum	MIC50
Cefepime	8	≥256	64	7
C/T	0.5	≥256	4	50
Meropenem	8	≥32	≥32	0
Amikacin	0.75	≥256	16	57
Fosfomycin	3	≥1,024	64	53a
WCK 5222	2	32	8	97b

^aMIC ≤ 64 mg/liter.

^bMIC ≤ 16 mg/liter, where the gradient diffusion strip contains cefepime and zidebactam in a 1:1 cefepime-zidebactam ratio; 29 of 30 (97%) strains were inhibited at cefepime concentrations of 16 mg/liter in the presence of 16 mg/liter zidebactam.

Among 146 antibiotic interactions assessed, the majority of combinations produced additive effects (Table 2). Antagonism was not observed. Rarely, the susceptibility of the β-lactam was apparently restored by AMK or FOF (Table 2, Fig. 1). Either the β-lactam or AMK or FOF inhibited growth below susceptibility breakpoints (FOF, ≤ 64 mg/liter) in 105/146 (72%) assessments. Clinicians must realize that in vitro synergy, or restoration of susceptibility in the GDS cross assay, may not translate to microbiological and/or clinical cure. There is, however, promising evidence in favor of synergy-guided antibiotic combination therapy (12).

TABLE 2.

Antibiotic interactions and restored β-lactam susceptibility^a

β-Lactam combination	No. (%) of isolates
	Synergy	Additivity	Indifference	RS	RS and synergy	RS and additivity
FEP+AMK	4 (14)	20 (71)	4 (14)	1 (4)	1 (100)	0 (0)
FEP+FOF	5 (18)	20 (71)	3 (11)	1 (4)	1 (100)	0 (0)
C/T+AMK	3 (20)	10 (67)	2 (13)	7 (47)	1 (14)	6 (86)
C/T+FOF	5 (33)	8 (53)	2 (13)	5 (33)	2 (40)	3 (60)
MEM+AMK	1 (3)	11 (37)	18 (60)	0 (0)
MEM+FOF	5 (17)	9 (30)	16 (53)	0 (0)

^aAMK, amikacin; C/T, ceftolozane-tazobactam; FEP, cefepime; FOF, fosfomycin; MEM, meropenem; RS, restored susceptibility. FEP combinations were assessed in 28 FEP-nonsusceptible isolates; C/T combinations were assessed in 15 C/T-nonsusceptible isolates; MEM combinations were assessed in all 30 CRP isolates.

FIG 1.

Example of restored ceftolozane-tazobactam (C/T) susceptibility in combination with amikacin (AMK). The C/T MIC alone was 8 mg/liter, but in combination with AMK, it was reduced to 3 mg/liter. The amikacin GDS MIC was 16 mg/liter prior to synergy testing.

While definitive treatment of P. aeruginosa with two antipseudomonal antibiotics also remains a controversial topic (13, 14), empirical coverage with two agents may be appropriate on a case-by-case basis. A recent U.S. surveillance study of over 1,200 P. aeruginosa isolates revealed an exceedingly high C/T susceptibility rate (>90%); in contrast, the susceptibility rate in the present study of 30 CRP strains was low (50%; Table 1). If these 30 strains are considered to be representative of CRP strains causing infections in regions burdened with increasing carbapenem resistance rates, an optimal empirical regimen would be represented by any combination that contained ≥1 antibiotic that inhibited growth at concentrations lower than an established susceptibility breakpoint (9), either when tested alone or in combination. This “best case scenario” among the 30 strains was C/T plus AMK (26/30, 87%), which was not as active as WCK 5222 alone, considering that 29/30 (97%) strains were inhibited at concentrations of 16 mg/liter, a potential breakpoint supported by in vivo data (5).

Against clinical CRP, WCK 5222 demonstrated remarkable in vitro potency and greater activity than currently available broad spectrum β-lactams, including ceftolozane-tazobactam. Coverage achieved with WCK 5222 alone was comparable to that achieved with C/T plus amikacin or fosfomycin. Taken together with the recently reported in vivo efficacy of WCK 5222 human-simulated exposures against MDR P. aeruginosa (5), our results suggest that WCK 5222, once approved as a single intravenous product, will be a future key player in empirical and definitive treatment of MDR and Carb^r P. aeruginosa infections. Clinical and in vivo studies are needed to further characterize and assess the efficacy of WCK 5222 monotherapy compared to antibiotic combination therapies.