In Vitro Activity of Ceftolozane-Tazobactam against Multidrug-Resistant Nonfermenting Gram-Negative Bacilli Isolated from Patients with Cystic Fibrosis

ABSTRACT

Ceftolozane-tazobactam was tested against 58 multidrug-resistant nonfermenting Gram-negative bacilli (35 Pseudomonas aeruginosa, 11 Achromobacter xylosoxydans, and 12 Stenotrophomonas maltophilia isolates) isolated from cystic fibrosis patients and was compared to ceftolozane alone, ceftazidime, meropenem, and piperacillin-tazobactam. Ceftolozane-tazobactam was the most active agent against P. aeruginosa but was inactive against A. xylosoxydans and S. maltophilia. In time-kill experiments, ceftolozane-tazobactam had complete bactericidal activity against 2/6 clinical isolates (33%).

TEXT

Nonfermenting Gram-negative bacilli (NF-GNB) are well-known pulmonary pathogens that colonize the airways of patients with cystic fibrosis (CF) (1, 2). Acute pulmonary exacerbations and chronic infections due to Pseudomonas aeruginosa, Achromobacter xylosoxidans, and Stenotrophomonas maltophilia may result in decreases in pulmonary function and high morbidity and mortality rates among CF patients (3–5). It was shown that multidrug-resistant (MDR) P. aeruginosa strains were associated with accelerated progression of CF and unfavorable outcomes (6).

Although the ceftolozane-tazobactam combination was tested previously in vitro against a large panel of MDR Enterobacteriaceae and P. aeruginosa isolates (7–10), there is little information concerning its activity against MDR strains isolated from CF patients. The activity of ceftolozane alone against P. aeruginosa isolates with high rates of β-lactam susceptibility from chronically infected CF patients was reported in two studies with encouraging results (11, 12). Only one study reported the activity of ceftolozane in combination with tazobactam, which is the currently marketed combination (Zerbaxa; Merck & Co., Inc.) (13). To our knowledge, ceftolozane-tazobactam has never been tested against MDR Achromobacter and S. maltophilia isolates from CF patients.

In this study, we measured the MIC and bactericidal activity of ceftolozane, alone and in combination with tazobactam, against various NF-GNB isolated from chronically infected CF patients that were highly resistant to most antipseudomonal β-lactams. Although tazobactam was not expected to enhance the activity of ceftolozane against MDR NF-GNB from CF patients, ceftolozane and ceftolozane-tazobactam were tested in parallel, in order to compare our results with those of other studies that evaluated ceftolozane alone and/or ceftolozane-tazobactam.

Bacterial strains and susceptibility results.

Fifty-eight nonduplicate clinical isolates of NF-GNB (35 P. aeruginosa, 11 A. xylosoxydans, and 12 S. maltophilia isolates) from 55 adult patients with CF were collected in two French teaching hospitals. All isolates were from respiratory tract samples and were identified by matrix-assisted laser desorption ionization–time of flight mass spectrometry (Bruker Daltonics, Bremen, Germany). All strains showed a high level of resistance to β-lactams, defined by resistance to ceftazidime determined with the disk diffusion method (inhibition zone diameter of <16 mm), according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) clinical breakpoints (http://www.eucast.org/clinical_breakpoints). MICs were determined using the agar dilution method, according to the 2015 recommendations of the Clinical and Laboratory Standards Institute (CLSI) (14). The MICs of the following antibiotics were determined: ceftolozane (MSD, Courbevoie, France), alone and in combination with tazobactam (Sigma-Aldrich, Lyon, France), piperacillin-tazobactam (Panpharma, Luitré, France), ceftazidime (Panpharma), and meropenem (Fresenius Kabi, Sèvres, France). Tazobactam was used at the fixed concentration of 4 μg/ml. The MICs of all antibiotics were determined for all strains in triplicate, after 24 h of incubation at 35 ± 2°C, as recommended (14). When bacterial growth was insufficient, incubation was extended to 48 h. P. aeruginosa ATCC 27853 and Klebsiella pneumoniae ATCC 700603 (an SHV-18-producing strain for the control of tazobactam combinations) were used as reference strains throughout this study.

Ceftolozane was the most active compound against P. aeruginosa isolates, with MIC₅₀ and MIC₉₀ values of 4 μg/ml and 128 μg/ml, respectively (Table 1). The combination with 4 μg/ml tazobactam did not improve the activity of ceftolozane against P. aeruginosa; this is not surprising, since CF isolates rarely produce acquired class A β-lactamases, which are the main targets for tazobactam (15, 16). The MIC₅₀ of ceftolozane-tazobactam was 5 doubling dilutions lower than those of ceftazidime and piperacillin-tazobactam and 2 doubling dilutions lower than that of meropenem. Considering the EUCAST clinical breakpoints, the highest rate of susceptibility (54%) was among the P. aeruginosa isolates with the ceftolozane-tazobactam combination (Table 1). Among 24 P. aeruginosa isolates that were nonsusceptible (i.e., intermediate or resistant) to all comparator antibiotics, ceftolozane-tazobactam remained active against 10/24 isolates (42%). Among the 16 P. aeruginosa isolates that were resistant to ceftolozane-tazobactam (MIC range, 8 to >128 μg/ml; median MIC, 8 μg/ml), piperacillin-tazobactam and meropenem were active against only one isolate each, while all 16 isolates were resistant to ceftazidime (MICs of ≥16 μg/ml) (data not shown). These results are consistent with previous studies of P. aeruginosa isolates from CF patients, which showed (i) greater activity of ceftolozane against strains for which ceftazidime MIC₅₀ values ranged from 1 to 8 μg/ml (11, 13) and (ii) a ceftolozane MIC₅₀ value 3 doubling dilutions lower than that of ceftazidime for strains with decreased susceptibility to ceftazidime (MIC₅₀ of 32 μg/ml) (12).

TABLE 1.

Susceptibility data for ceftolozane-tazobactam and β-lactam comparators in a collection of 58 nonfermenting Gram-negative bacilli isolated from cystic fibrosis patients

Species and antimicrobial agent(s)	MIC (μg/ml)			EUCAST clinical breakpoint (μg/ml)	No. of strains			Proportion of susceptible strains (%)
	MIC range	MIC50	MIC90		Resistant	Intermediate	Susceptible
All isolates (n = 58)
Ceftolozane	0.5 to >128	16	>128
Ceftolozane + tazobactam	0.5 to >128	16	>128
Ceftazidime	4 to >128	128	>128
Piperacillin + tazobactam	1 to >128	64	>128
Meropenem	0.5 to >128	32	>128
Pseudomonas aeruginosa (n = 35)
Ceftolozane	0.5 to >128	4	128
Ceftolozane + tazobactama	0.5 to >128	4	128	≤4 to >4	16		19	54
Ceftazidime	4 to >128	128	>128	≤8 to >8	31		4	11
Piperacillin + tazobactam	2 to >128	128	>128	≤16 to >16	26		9	26
Meropenem	1 to >128	16	128	≤2 to >8	26	6	3	9
Stenotrophomonas maltophilia (n = 12)
Ceftolozane	16 to >128	64	>128
Ceftolozane + tazobactam	16 to >128	64	>128	≤4 to >4	12		0	0
Ceftazidime	16 to >128	128	>128	≤8 to >16	11	1	0	0
Piperacillin + tazobactam	8 to >128	128	>128	≤4 to >16	11	0	1	8
Meropenem	4 to >128	>128	>128	≤2 to >8	10	2	0	0
Achromobacter xylosoxydans (n = 11)
Ceftolozane	>128	>128	>128
Ceftolozane + tazobactam	>128	>128	>128	≤4 to >4	11		0	0
Ceftazidime	32 to >128	>128	>128	≤4 to >8	11	0	0	0
Piperacillin + tazobactam	1 to 64	16	32	≤4 to >16	3	6	2	18
Meropenem	0.5 to >128	32	128	≤2 to >8	9	1	1	9

^aWhen CLSI clinical breakpoints were applied, the numbers of P. aeruginosa isolates resistant, intermediate, and susceptible to ceftolozane-tazobactam were 8, 8, and 19, respectively. The proportion of susceptible strains remained unchanged (54%).

Piperacillin-tazobactam was the most active antibiotic against A. xylosoxydans, followed by meropenem, while ceftolozane-tazobactam and ceftazidime showed no activity (Table 1). None of the β-lactams tested was active against MDR S. maltophilia isolates (Table 1), which is consistent with previous results obtained for stains isolated from non-CF patients (9).

Bactericidal activity of ceftolozane-tazobactam.

Considering the lower MICs of ceftolozane-tazobactam, in comparison with other antipseudomonal β-lactams, for P. aeruginosa, time-kill assays were performed to evaluate the bactericidal activity of ceftolozane-tazobactam against various P. aeruginosa isolates. The bactericidal activity of ceftolozane in combination with 4 μg/ml tazobactam was determined as described previously (17). Briefly, a bacterial suspension incubated overnight at 35°C in brain heart infusion broth (BHI-B) was diluted 1/1,000 in fresh BHI-B and incubated for 2 h at 35°C, with agitation, to reach a bacterial density of 10⁵ to 10⁶ CFU/ml. Antibiotics were added at concentrations corresponding to 4 or 8 times their MIC values, and cultures were incubated for 24 h at 35°C, with agitation. CFU counts were performed at 0, 2, 6, 8, and 24 h, using adequate dilutions plated on Mueller-Hinton (MH) agar. Antibiotic-free cultures were included as growth controls in all assays. Bactericidal activities of antipseudomonal β-lactam antibiotics (ceftolozane alone, ceftazidime, piperacillin-tazobactam, and meropenem) were compared with those of ceftolozane-tazobactam against the reference strain P. aeruginosa ATCC 27853. A complete bactericidal effect was defined as a ≥3-log₁₀ decrease from the initial bacterial colony count. For the reference strain P. aeruginosa ATCC 27853 (ceftolozane-tazobactam MIC of 0.5 μg/ml), a complete bactericidal effect (defined as a ≥3-log₁₀ decrease from the initial colony count) was observed 6 h after the addition of ceftolozane-tazobactam at both 4× MIC and 8× MIC (Fig. 1). Ceftolozane alone showed similar results, indicating that tazobactam did not improve the bactericidal activity of ceftolozane. Ceftazidime had comparable bactericidal activity. Meropenem achieved complete bactericidal activity more quickly, with a 3-log₁₀ reduction at 8× MIC after 4 h, compared to approximately 6 to 7 h for the other compounds (Fig. 1B). No regrowth was observed within 24 h in the presence of ceftolozane, ceftolozane-tazobactam, or meropenem, while an increase in colony counts was observed with piperacillin-tazobactam, especially at 4× MIC.

FIG 1.

Time-kill curves for P. aeruginosa ATCC 27853 exposed to antimicrobial agents at 4× MIC (A) or 8× MIC (B). Antibiotics were added to the medium after 2 h of incubation. □, growth control; ▲, ceftolozane alone; △, ceftolozane-tazobactam; ●, ceftazidime; ■, piperacillin-tazobactam; ◆, meropenem.

The bactericidal activity was also studied with six P. aeruginosa clinical isolates for which the MICs of ceftolozane-tazobactam were 4 μg/ml (upper limit of susceptibility [three strains]) or 8 μg/ml (resistance [three strains]) (Fig. 2). The highest concentration of ceftolozane-tazobactam tested for these strains (64 μg/ml in the case of the three resistant strains) remained in the ranges achievable in human serum, according to previous pharmacokinetic studies conducted with standard dosing regimens of 1,000 mg/500 mg three times a day (18, 19). For all six strains, a >1.5-log₁₀ decrease in the number of CFU per milliliter was observed after 6 h of incubation in the presence of antibiotic (Fig. 2). However, complete bactericidal activity was achieved for only two strains (Fig. 2A and D) using 8× MIC of ceftolozane-tazobactam; only partial bactericidal activity was achieved for all isolates at 4× MIC. Regrowth was observed for all P. aeruginosa isolates after 24 h of incubation.

FIG 2.

Time-kill curves for clinical P. aeruginosa isolates exposed to ceftolozane-tazobactam. Antibiotics were added to the medium after 2 h of incubation. The MICs, for three isolates each, were 4 μg/ml (A to C) or 8 μg/ml (D to F). □, growth control; ▲, 4× MIC; ●, 8× MIC.

Similar regrowth was observed with other combinations of β-lactams, such as piperacillin-tazobactam and ceftazidime-avibactam, in previous studies (20) and also with piperacillin-tazobactam at 4× MIC against P. aeruginosa ATCC 27853 in the present study. As the wild-type P. aeruginosa ATCC 27853 showed no regrowth when exposed to ceftolozane-tazobactam (Fig. 1), an inoculum effect due to AmpC overexpression cannot be excluded for clinical isolates. The amplification of a ceftolozane-resistant subpopulation was previously suggested in time-kill assays performed with Escherichia coli (21). Therefore, 24-h colonies obtained after regrowth were plated on brain heart infusion (BHI) agar with increasing concentrations of ceftolozane-tazobactam. Growth at concentrations higher than the initial MIC was observed for 3/6 isolates (50%), for which the initial and final MICs were 4 and 32 mg/liter, 4 and >64 mg/liter, and 8 and >64 mg/liter, respectively. This finding reveals the in vitro selection of ceftolozane-tazobactam-resistant P. aeruginosa populations, suggesting that resistant strains could easily emerge in everyday clinical practice, especially among CF patients, who are usually subjected to long courses of antibiotic treatments. Appropriate and thoughtful use of this antibiotic should be considered by prescribers.

In conclusion, based on MIC determinations as well as time-kill assays, the ceftolozane-tazobactam combination appears to be a major antipseudomonal drug, especially for CF patients, who are often chronic carriers of P. aeruginosa strains that are highly resistant to other antipseudomonal β-lactams. The ceftolozane-tazobactam activity is insufficient, however, against other CF NF-GNB, such as S. maltophilia and A. xylosoxidans.