Pharmacodynamic activity of ertapenem versus multidrug-resistant genotypically characterized extended-spectrum β-lactamase-producing Escherichia coli using an in vitro model

Abstract

Background

This study assessed the pharmacodynamic activity of ertapenem against multidrug-resistant (MDR) genotypically characterized extended-spectrum β-lactamase (ESBL)-producing Escherichia coli using an in vitro model.

Methods

Six ESBL-producing E. coli with the CTX-M-15 genotype were studied. All six strains were MDR (defined as resistance to third-generation cephalosporins and ≥two other unrelated antimicrobial classes). The in vitro pharmacodynamic model was inoculated with 1 × 10⁶ cfu/mL, and ertapenem was dosed once daily at 0 and 24 h to simulate free (f) C_max and t_1/2 obtained after a standard 1 g intravenous once-daily dose in healthy volunteers (fC_max, 15 mg/L; t_1/2, 4 h). Sampling was performed over 48 h to assess viable growth and resistance selection.

Results

Ertapenem T_>MIC ≥98% (ertapenem MICs ≤0.25 mg/L) resulted in bactericidal (≥3 log₁₀ killing) activity at 6, 12, 24 and 48 h against all strains. Eradication of organisms from the in vitro model (below the level of detection) occurred at 2 h followed by slow regrowth of the majority of the strains (5 of 6) over 12, 24 and 48 h time points. Despite limited regrowth, ertapenem achieved a bactericidal effect against all strains (all time points) over the 48 h study period.

Conclusions

Ertapenem was rapidly bactericidal (in ∼2 h) against MDR ESBL (CTX-M-15)-producing E. coli (ertapenem MICs ≤0.25 mg/L) when simulating free drug after 1 g intravenous once-daily dosing. This bactericidal activity was maintained over the 48 h experimental period with only minor regrowth, which was not associated with MIC increase from baseline.

resistance, antimicrobials, ESBLs, E. coli

Introduction

Escherichia coli is a common pathogen linked with community-associated as well as nosocomial infections. The emergence and evolving spread of extended-spectrum β-lactamase (ESBL)-producing E. coli in the community, extended-care facilities and hospital settings has been well documented.^1,2 ESBL-producing E. coli are frequently multidrug-resistant (MDR; defined as resistant to third-generation cephalosporins and ≥two other unrelated antimicrobial classes).^2,3 Carbapenems such as imipenem/cilastatin and meropenem are recognized as the drugs of choice for seriously ill patients with ESBL E. coli infections.^4,5 However, little data are available regarding the pharmacodynamic outcomes with ertapenem against MDR ESBL-producing E. coli.

Ertapenem is a new carbapenem developed to address the pharmacokinetic shortcomings (short half-life) of imipenem and meropenem.⁵ Ertapenem shares similar structural features with meropenem including its stability to dehydropeptidase-I (DHP-I), allowing it to be administered without a DPH-I inhibitor. Ertapenem, like imipenem and meropenem, demonstrates broad-spectrum antimicrobial activity against many Gram-positive and Gram-negative aerobes and anaerobes, and is resistant to nearly all β-lactamases including ESBLs and AmpCs.⁵ It, however, differs from both imipenem and meropenem in demonstrating limited activity against Enterococcus spp., Pseudomonas aeruginosa and other non-fermentative Gram-negative bacteria commonly associated with nosocomial infections. Extensive protein binding of ertapenem extends the half-life and allows for once-daily dosing.⁵ Clinical trials have demonstrated that ertapenem has equivalent efficacy and safety compared with ceftriaxone and piperacillin/tazobactam against a variety of community-acquired infections.

The purpose of this study was to assess the pharmacodynamic activity of ertapenem against MDR genotypically characterized ESBL-producing E. coli using an in vitro pharmacodynamic model.

Materials and methods

Bacterial strains and culture conditions

The six ESBL E. coli isolates were obtained from the Canadian National Intensive Care Unit (CAN-ICU) study (www.can-r.ca) (T. Baudry, K. A. Nichol, M. DeCorby, P. Lagacé-Wiens, E. Becker, M. Mulvey, D. J. Hoban and G. G. Zhanel, unpublished results). In the CAN-ICU study, any E. coli with a ceftriaxone MIC ≥1 mg/L was identified as a putative ESBL. Putative ESBL phenotypes were confirmed by the disc diffusion method as described by the CLSI. E. coli ATCC 25922 and Klebsiella pneumoniae ATCC 700603 were the control strains. The six ESBL E. coli strains from CAN-ICU were chosen for this study because of their geographic variability (obtained from western, central and eastern Canada) as well as genetic heterogeneity (pulsed field gel electrophoresis patterns, not shown). Genotypic characterization of ESBLs was performed by PCR and sequencing of bla_SHV, bla_TEM, bla_CTX, bla_OXA and bla_VEB genes as previously described.¹ A BLAST search of the DNA sequence was conducted to determine the specific ESBL genotype. PCR and sequencing revealed that all six ESBL E. coli were bla_CTX-M-15. Four out of six (67%) strains were also bla_oxa-1 but none had plasmid-mediated bla_AmpC genes (T. Baudry, K. A. Nichol, M. DeCorby, P. Lagacé-Wiens, E. Becker, M. Mulvey, D. J. Hoban and G. G. Zhanel, unpublished results). All six ESBL E. coli strains were chosen because they had a MDR (defined as resistance to third-generation cephalosporins and ≥two other unrelated antimicrobial classes) phenotype. Strains were chosen with varying ertapenem MICs and at least one strain was chosen from the pool of ESBL E. coli that had the highest ertapenem MIC.

For the pharmacodynamic studies, logarithmic phase cultures with turbidities equivalent to that of a 0.5 McFarland standard (1 × 10⁸ cfu/mL) in cation-supplemented Mueller–Hinton broth were prepared as previously described.⁶ Viable bacterial counts consistently yielded a starting inoculum of ∼1 × 10⁶ cfu/mL. A growth control was included in every experiment. Growth controls peaked at ∼1 × 10⁹ cfu/mL and were maintained over the 48 h experiment.

Antibiotic preparations and susceptibility testing

Antibiotic agents were obtained as laboratory-grade powders from their respective manufacturers (Ertapenem, Merck, Montreal, Quebec, Canada). Stock solutions were made according to the CLSI (formerly the NCCLS) M7-A6 method. MICs were determined by the CLSI-approved broth microdilution method. All MICs were performed in triplicate on separate days.

Pharmacokinetics of ertapenem in the in vitro pharmacodynamic model

Experiments were performed simulating peak serum concentrations (C_max) and AUC₂₄ of ertapenem, achieved in human serum after standard intravenous doses (1 g of ertapenem once daily) (Table 1).^6–8 Protein-free (f; unbound) serum concentrations were simulated using known protein binding fractions (ertapenem ∼90%).^6–8 Ertapenem clearance was simulated using a reported serum half-life of 4 h.⁵ The pharmacokinetics of ertapenem were evaluated by dosing using standard doses in the central compartment and sampling from this compartment at 0, 1, 2, 4, 6, 12, 18, 24, 36 and 48 h. Ertapenem concentrations were determined in quadruplicate using Bacillus subtilis ATCC 6633 as the test organism with a lower limit of quantification of 0.25 mg/L as previously described.⁶ The correlation coefficient of this assay was 0.84. The intra-day and inter-day coefficients of variation were 3.2% to 5.2%, and 2.8% to 5.1%, respectively. The fAUC₂₄ (mg·h/L) for ertapenem was calculated using the trapezoidal rule.⁶ The fAUC₂₄/MIC was calculated for ertapenem against the specific E. coli studied.

Table 1

Ertapenem pharmacodynamic parameters simulated

Strain	Ertapenem MIC (mg/L)	T>MIC, h (%)	fCmax/MIC	fAUC24/MIC
59096	0.06	24 (100)	230	1100
62188	0.06	24 (100)	230	1100
63106	0.06	24 (100)	230	1100
61567	0.12	24 (100)	115	550
64197	0.12	24 (100)	115	550
64771	0.25	23.6 (98)	55.2	264

In vitro pharmacodynamic model/pharmacodynamic experiments

The in vitro pharmacodynamic model used in this study has been previously described.⁶ Logarithmic phase cultures were diluted into fresh cation-supplemented Mueller–Hinton broth to achieve a final inoculum of ∼1 × 10⁶ cfu/mL. Only free (protein unbound) serum concentrations were simulated. Pharmacodynamic experiments were performed in duplicate (on separate days) in ambient air at 37°C. At 0, 1, 2, 4, 6, 12, 18, 24, 36 and 48 h, samples were removed from the central compartment and viable bacterial counts performed by plating serial 10-fold dilutions onto cation-supplemented Mueller–Hinton agar as previously described.⁶ The lowest dilution plated was 0.1 mL of undiluted sample and the lowest level of detection was 200 cfu/mL (20 colonies of 0.1 mL undiluted sample). Antibiotic carryover was minimized by diluting samples withdrawn from the model or by repeated washing and centrifugation. No difference in antibiotic carry-over was observed between dilution and washing. Measurement of antibacterial effects was assessed as log₁₀ changes in bacterial counts at 6, 12, 24 and 48 h with respect to time 0.

Results

All six ESBL-producing E. coli were genotyped as CTX-M-15 and were isolated from the following clinical sites: blood (n = 3), respiratory tract (n = 1), urine (n = 1) and wound (n = 1). All six strains were MDR (defined as resistance to third-generation cephalosporins and ≥two other unrelated class of antimicrobial). Resistance rates of the six MDR ESBL E. coli were: 100% (6/6) to ceftriaxone, 100% (6/6) to ciprofloxacin, 67% (4/6) to gentamicin, 67% (4/6) to trimethoprim/sulfamethoxazole, 67% (4/6) to doxycycline and 33% (2/6) to piperacillin/tazobactam. All isolates were susceptible to tigecycline and ertapenem. Ertapenem MICs (mg/L) [number of strains] studied were: 0.06 [3], 0.12 [2] and 0.25 [1]. Strain 64771 displayed the highest ertapenem MIC (0.25 mg/L) of the six strains studied and displayed the highest ertapenem MIC of all 536 E. coli collected from the CAN-ICU study.

The simulated and achieved pharmacokinetic profiles of ertapenem in the central compartment of the pharmacodynamic model were within 15% to 20% of simulated pharmacokinetic values. Achieved ertapenem pharmacokinetics were fC_max 13.8 ± 1.8 mg/L (simulated fC_max 15.0 mg/L), t_1/2 3.7 ± 0.5 h (simulated t_1/2 4 h) and fAUC₂₄ 66.0 ± 7.7 mg·h/L (simulated fAUC₂₄ 57 mg·h/L). The achieved ertapenem pharmacodynamics were fC_max/MIC 55.2–230, fAUC₂₄/MIC 264–1100 and T_>MIC 98% to 100% (Table 1).

The pharmacodynamic activity of ertapenem against MDR ESBL-producing E. coli, simulating free serum concentrations, is displayed in Table 2. Ertapenem T_>MIC ≥98% (ertapenem MICs ≤0.25 mg/L) resulted in bactericidal (≥3 log₁₀ killing) activity at 6, 12, 24 and 48 h against all six strains. With all six strains, rapid eradication of the organisms from the in vitro model (below the level of detection) occurred by the 2 h mark (data not shown), which was followed by slow regrowth of the majority of the strains (5 of 6) over 12, 24 and 48 h time points (despite re-dosing ertapenem at 24 h) (Table 2). Despite limited regrowth, ertapenem achieved a bactericidal effect against all strains at all tested time points over the 48 h study period (Table 2). The observed MICs for E. coli of ertapenem studied in the in vitro model did not change during the 48 h period, even for strains where minor regrowth occurred (Table 2).

Table 2

Ertapenem killing of ESBL E. coli simulating free serum concentrations

		Log10 killing at 6, 12, 24 and 48 h, respectivelya
Strain [ertapenem MIC (mg/L)]	T>MIC (%)	6	12	24	48
59096 (0.06)	100	≥4.0	≥4.0	4.0	3.8 ± 0.4 (0.06)b
62188 (0.06)	100	≥4.0	3.5 ± 0.3 (0.06)	3.0 ± 0.5	3.0 ± 0.4 (0.06)
63106 (0.06)	100	≥4.0	3.5 ± 0.3 (0.06)	3.3 ± 0.3	3.0 ± 0.2 (0.06)
61567 (0.12)	100	≥4.0	3.8 ± 0.4 (0.12)	3.2 ± 0.3	3.1 ± 0.2 (0.12)
64197 (0.12)	100	≥4.0	3.5 ± 0.5 (0.12)	3.3 ± 0.4	3.1 ± 0.5 (0.12)
64771 (0.25)	98	≥4.0	3.2 ± 0.3 (0.25)	3.0 ± 0.4	3.3 ± 0.6 (0.25)

^aGrowth reduction relative to initial inoculum.

^bMIC determined by Etest (on freshly isolated colonies).

Discussion

ESBL-producing E. coli are frequently MDR and limited options exist to treat these organisms.^2,3 Thus, we used an in vitro pharmacodynamic model to simulate ertapenem pharmacokinetic (fC_max and fAUC₂₄) and pharmacodynamic parameters against MDR genotypically characterized ESBL-producing E. coli.⁵ The six strains studied were selected because of geographic variability, genetic heterogeneity and displayed varying ertapenem MICs, and at least one strain was chosen from the pool of ESBL E. coli that had the highest ertapenem MIC (0.25 mg/L). Other investigators have reported that ESBL E. coli demonstrate ertapenem MIC₅₀ and MIC₉₀ ranging from ≤0.03 to 0.06 and 0.125 to 0.25 mg/L, respectively.^9–11 ESBL E. coli which demonstrate ertapenem MICs >0.25 mg/L are not common.^9–11 A weakness of our study was the fact that we were unable to collect and study ESBL E. coli with ertapenem MIC of >0.25 mg/L; however, we have been unable to locate such a clinical isolate from patients across Canada.

This study showed that ertapenem was rapidly (∼2 h) bactericidal against all six MDR ESBL E. coli studied and maintained the bactericidal activity over the 48 h interval, when simulating exposure of free drug after 1 g intravenous once-daily dosing (Table 2). Pharmacokinetically, ertapenem although highly protein bound (∼90%),⁵ achieved a prolonged T_>MIC (≥98%) of free (unbound) drug resulting in bactericidal activity against MDR ESBL E. coli with ertapenem MIC ≤0.25 mg/L (Table 2). Our results are in agreement with DeRyke et al.¹² who reported ∼2 log₁₀ killing of ESBL E. coli (ertapenem MICs <0.5 mg/L, fT_>MIC ≥53%) after 24 h of therapy with ertapenem using a murine neutropenic thigh infection model. These authors also achieved ∼2 log₁₀ killing against ESBL E. coli with ertapenem MICs 0.5–1.5 mg/L (fT_>MIC 23% to 40%).¹² Against strains with ertapenem MIC ≥2 mg/L (fT_>MIC ≤20%) little-to-no killing occurred.¹² These data suggest that ertapenem is a reasonable option for the treatment of infections caused by MDR ESBL E. coli with MIC <2 mg/L. Prudent use of ertapenem and all carbapenems is, however, required as a highly carbapenem-resistant (ertapenem MIC >256 mg/L, imipenem/cilastatin 8 mg/L and meropenem 8 mg/L) E. coli, which has been documented in a patient who received 10 days if on imipenem/cilastatin treatment.¹³ Ertapenem resistance was explained by both an outer membrane defect as well as an ESBL CTX-M-2 enzyme.¹³

In conclusion, ertapenem was rapidly bactericidal (in ∼2 h) against MDR ESBL (CTX-M-15)-producing E. coli (ertapenem MICs ≤0.25 mg/L) when simulating free drug after 1 g intravenous once-daily dosing. This bactericidal activity was maintained over the 48 h experimental period with only minor regrowth which was not associated with MIC increase from baseline.

Funding

This study was supported in part by the University of Manitoba and Health Sciences Center.

Transparency declarations

P. B. received funding from a University of Manitoba fellowship. All other authors: none to declare.