Fosfomycin Enhances the Activity of Daptomycin against Vancomycin-Resistant Enterococci in an In Vitro Pharmacokinetic-Pharmacodynamic Model

Abstract

Daptomycin (DAP) is being used more frequently to treat infections caused by vancomycin-resistant enterococcus (VRE). DAP tends to be less active against enterococci than staphylococci and may require high doses or combination therapy to be bactericidal. Fosfomycin (FOF) has activity against VRE and has demonstrated synergistic bactericidal activity with DAP in vitro. The objective of this study was to evaluate the activity of DAP alone and in combination with FOF against VRE in an in vitro pharmacokinetic/pharmacodynamic (PK/PD) model. The activity of DAP at 8 and 12 mg/kg of body weight/day (DAP 8 and DAP 12, respectively) and FOF of 40 mg/kg intravenously every 8 h, alone and in combination, were evaluated against 2 vancomycin-resistant Enterococcus faecium strains (8019 and 5938) and 2 vancomycin-resistant E. faecalis strains (V583 and R7302) in an in vitro PK/PD model over 72 h. Cell surface charge in the presence and absence of FOF was evaluated by zeta potential analysis. Daptomycin-boron-dipyrromethene (bodipy) binding was assessed by fluorescence microscopy. The addition of FOF to DAP 8 and DAP 12 resulted in significantly increased killing over DAP alone at 72 h for 8019, V583, and R7302 (P < 0.05). Therapeutic enhancement was observed with DAP 12 plus FOF against 8019, V583, and R7302. Cell surface charge became more negative after exposure to FOF by ∼2 to 8mV in all 4 strains. Daptomycin-bodipy binding increased by 2.6 times in the presence of fosfomycin (P < 0.0001). The combination of DAP plus FOF may provide improved killing against VRE (including DAP-resistant strains) through modulation of cell surface charge. Further studies to clarify the role of intravenous FOF are warranted.

INTRODUCTION

Daptomycin (DAP) is a cationic cyclic lipopeptide demonstrating rapid bactericidal activity against Gram-positive bacteria, including vancomycin-resistant Enterococcus (VRE), through disruption of electrochemical membrane potential (1). Fosfomycin (FOF) is a phosphonic acid derivative that inhibits peptidoglycan synthesis in both Gram-negative and Gram-positive organisms. It displays activity against methicillin-resistant Staphylococcus aureus (MRSA), multidrug-resistant Streptococcus pneumoniae, Enterococcus spp., including vancomycin-resistant strains, and Gram-negative bacteria, including extended-spectrum β-lactamase (ESBL)-producing Enterobacter (2–4). The following MIC interpretive criteria have been established by the Clinical and Laboratory Standards Institute (CLSI) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST): Enterococcus faecalis (urinary isolates only) at ≤64, 128, and ≥256 mg/liter as susceptible (S), intermediate (I), and resistant (R), respectively, by CLSI; Staphylococcus spp. at ≤32 and ≥32 mg/liter as S and R by EUCAST (2, 5; http://www.eucast.org). It primarily undergoes renal elimination and is indicated for treatment of uncomplicated urinary tract infection (UTI) caused by Escherichia coli and Enterococcus faecalis (2, 6). Although an oral preparation approved for urinary tract infections is the only fosfomycin product available in the United States, the intravenous formulation has been used outside the United States for treatment of severe systemic infections (7). Fosfomycin doses of 3 g orally have been well established to treat UTIs, and bloodstream infections have been treated with fosfomycin at 50 mg/kg of body weight three to four times daily in combination with a β-lactam agent (4, 8, 9). Other studies have utilized fosfomycin doses of 2 g every 6 or 8 h (10, 11). Developed in 1969, fosfomycin has drawn increasing attention for use in multidrug-resistant bacterial infections as combination therapy due to the increase in antimicrobial resistance and lack of novel antimicrobial development (6, 12, 13). While most of the data on fosfomycin-based antimicrobial combinations come from small in vitro studies, clinical trials are ongoing in Europe to evaluate the combination of daptomycin and intravenous fosfomycin for treatment of MRSA bacteremia (11).

VRE has been associated with increased clinical failure rates, and reports of clinical failure or resistance to newer antimicrobial agents, including daptomycin, are emerging (14–19). One large, retrospective cohort study found 2% of isolates developed nonsusceptibility to daptomycin, even when doses of >6 mg/kg/day were administered (20). Daptomycin and fosfomycin appear to be a promising combination for eradication of resistant Gram-positive infections. Several in vitro studies have demonstrated synergism. One study evaluated clinical strains by checkerboard and subsequently time-kill methods and found synergism against Enterococcus spp. (21). An additional study demonstrated significantly improved killing in vitro when combined with daptomycin against clinical isolates of vancomycin-resistant E. faecium than with either agent alone (22).

The mechanism of synergy between these agents is unknown. Synergy between daptomycin and beta-lactams is believed to be mediated through beta-lactam-induced alterations in surface charge that facilitate an increase in daptomycin binding, leading to enhanced membrane depolarization (23–28). Since fosfomycin also disrupts cell wall synthesis, it could be theorized that there is a similar mechanism that drives the synergy that has been observed with daptomycin and fosfomycin.

The purpose of this study was to evaluate the potential for synergistic effects of fosfomycin in combination with high-dose daptomycin against strains of vancomycin-resistant E. faecium and E. faecalis in an in vitro pharmacokinetic/pharmacodynamic (PK/PD) model simulating clinically relevant drug exposures.

(This study was presented in part at the 52nd Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, 2012, and the 23rd European Congress of Clinical Microbiology, Berlin, Germany, 2013.)

MATERIALS AND METHODS

Bacterial strains.

Four vancomycin-resistant enterococcus strains were evaluated: a clinical isogenic E. faecium strain pair (1 daptomycin-resistant [5939] and 1 daptomycin-susceptible [8019] strain) and 2 daptomycin-susceptible E. faecalis strains (V583 and R7203).

Antimicrobials and media.

DAP and FOF analytical powder were commercially purchased (Sigma-Aldrich, St. Louis, MO). Due to the calcium-dependent mechanism of DAP, Mueller-Hinton broth II supplemented to contain 50 μg/ml calcium and 12.5 μg/ml magnesium (SMHB) (Difco, Detroit, MI) was used for PK/PD models with DAP and FOF. For FOF-containing simulations, SMHB was supplemented with 25 μg/ml glucose-6-phosphate (G-6-P; Sigma-Aldrich, St. Louis, MO), which is a necessary cofactor for active transmembrane transport of FOF. Colony counts were determined using brain heart infusion agar (BHIA; Difco, Detroit, MI) plates. BHIA supplemented with 50 μg/ml of calcium or 25 μg/ml of G-6-P was used for antimicrobial resistance screening plates containing DAP or FOF, respectively. Mueller-Hinton agar (MHA) supplemented with 25 μg/ml of G-6-P was used for FOF bioassays.

Susceptibility testing.

MICs of DAP and FOF were determined in duplicate by Etest methodology at ∼1.5 × 10⁸ CFU/ml according to the manufacturer's recommendations. All samples were incubated at 37°C for 18 to 24 h. Development of nonsusceptibility was evaluated at 72 h for all simulated regimens. Samples (100 μl) were plated on FOF- or DAP-containing agar plates at 3× the MIC.

In vitro PK/PD model.

An in vitro, one-compartment PK/PD model with a 250-ml capacity and inflow and outflow ports was used. The apparatus was prefilled with medium, and antimicrobials were administered as boluses over a 72-h time period. Prior to each experiment, bacterial lawns from an overnight growth on BHI were suspended and added to each model to obtain a starting inoculum of ∼10⁸ CFU/ml. The model apparatus was maintained at 37°C throughout the experiment, and a magnetic stir bar was placed in the medium for thorough mixing of the drug in the model. Fresh medium was continuously supplied and removed from the compartment along with the drug via a peristaltic pump (Masterflex; Cole-Parmer Instrument Company, Chicago, IL) at an appropriate rate to simulate the average human half-lives of the antimicrobials. A total of 5 simulated regimens were evaluated on each isolate for 72 h: daptomycin at 8 mg/kg (termed DAP 8) every 24 h (q24h) (maximum concentration of free drug in serum [fC_max], 9.86 mg/liter; average half-life [t_1/2], 8 h; protein binding, 92%) (29), DAP at 8 mg/kg q24h plus FOF at 40 mg/kg q8h (fC_max, 260 mg/liter; average t_1/2, 2.3 h) (30), daptomycin at 12 mg/kg (DAP 12) q24h (fC_max, 14.7 mg/liter) (29), DAP at 12 mg/kg q 24 h plus FOF at 40 mg/kg q8h, and FOF at 40 mg/kg q8h. A growth control was also run for each isolate to verify isolate fitness. Models were performed in duplicate to ensure reproducibility. Supplemental DAP was added at an appropriate rate to FOF combination models to compensate for the higher flow rate required to simulate FOF clearance (31).

Pharmacodynamic analysis.

Samples from each model were collected at 0, 4, 8, 24, 32, 48, 56, and 72 h and diluted in 0.9% saline. Colony counts were determined by spiral plating appropriate dilutions using an automatic spiral plater (WASP; DW Scientific, West Yorkshire, England) to enumerate CFU/ml and avoid antibiotic carryover. Colonies were counted using an automated colony counter (ProtoCOL; Synoptics Limited, Frederick, MD). If the anticipated dilution was near the MIC, then vacuum filtration was also used to avoid antibiotic carryover. When vacuum filtration was used, samples were washed through a 0.45-μm filter with normal saline to remove the antimicrobial agent. For both methods, plates were incubated at 37°C for 18 to 24 h, at which time colony counts were performed. These methods have a lower limit of reliable detection of 1 log₁₀ CFU/ml.

The total reduction in log₁₀ CFU/ml over 72 h was determined by plotting time-kill curves based on the number of remaining organisms over the time period. Bactericidal activity (99.9% kill) and bacteriostatic activity were defined as a ≥3-log₁₀ CFU/ml or a <3-log₁₀ CFU/ml reduction, respectively, in colony count from the initial inoculum. Inactivity was defined as no observed reductions in initial inocula. The time to achieve a 99.9% bacterial load reduction was determined by linear regression (if r² ≥ 0.95) or visual inspection. The effects of the antimicrobial combinations were interpreted as follows. Enhancement of activity was defined as an increase in kill of ≥2-log₁₀ CFU/ml by the combination of antimicrobials versus the most active single agent of that combination. Improvement was defined as a 1- to 2-log₁₀ CFU/ml increase in kill compared to the most active single agent, while combinations that result in a ≥1-log₁₀ CFU/ml bacterial growth compared to the least active single agent were considered antagonistic (32, 33).

Pharmacokinetic analysis.

Pharmacokinetic samples were obtained through the injection port of each model at 0, 1, 2, 4, 8, 24, 32, 48, 56, and 72 h for verification of target antibiotic concentrations. All samples were stored at −70°C until ready for analysis. FOF concentrations were determined by bioassay using Escherichia coli (ATCC 25922). Blank 1/4-inch disks were spotted with 10 μl of standard concentrations or samples. Each standard was tested in duplicate by placing the disk on agar plates (MHA) inoculated with a 0.5 McFarland suspension of the test organism. This assay demonstrated an intraday coefficient of variance of less than 10% for all standards. Concentrations of daptomycin were determined using a validated high-performance liquid chromatography assay (34–36). This assay has demonstrated an interday coefficient of variation between 0.6 and 7.3% for all standards. The free antimicrobial peak concentrations, half-lives, and free 24-h area under the concentration-time curve (fAUC_{0–24 h}) were determined using PK Analyst software (version 1.10; MicroMath Scientific Software, Salt Lake City, UT). The trapezoidal method was utilized to calculate fAUC_{0–24 h}.

Zeta potential analysis.

Changes in cell surface charge in the presence and absence of subinhibitory concentrations of FOF were measured using a zeta potential analyzer. Briefly, cells were grown overnight in 5 ml of MHB supplemented with 25 mg/liter of G-6-P and 0.5× the MIC of FOF. Cell suspensions were adjusted to an optical density at 600 nm (OD₆₀₀) of 0.5 and then harvested by centrifugation. Cell pellets were washed 3 times and resuspended in sterile water. Zeta potential was measured using a Brookhaven Instruments ZetaPlus zeta potential analyzer. Results are averages from 10 measurements per organism per exposure. Experiments were conducted on 3 different days.

Daptomycin-bodipy binding studies.

Daptomycin binding was assessed by fluorescence microscopy in the presence and absence of fosfomycin and G-6-P as previously described (24). Briefly, cells were grown to an optical density at 600 nm (OD₆₀₀) of 0.6, grown for an additional 1 h with or without G-6-P (25 mg/liter) and G-6-P plus FOF (260 mg/liter), and then incubated with 16 mg/liter daptomycin-boron-dipyrromethene (bodipy) for 10 min. Cells were then washed three times in medium to remove unincorporated label, stained with 1 mg/liter 4′,6-diamidino-2-phenylindole (DAPI), and imaged as previously described prior to fluorescence quantification.

Statistical analysis.

Changes in CFU/ml at 24, 48, and 72 h were compared by one-way analysis of variance with Tukey's post hoc test. Zeta potential values and fluorescence intensity of daptomycin-bodipy binding were compared by student's t test. A P value of ≤0.05 was considered significant. All statistical analyses were performed using SPSS Statistical Software (release 20.0; SPSS, Inc., Chicago, IL).

RESULTS

Susceptibility testing and resistance.

Organism MICs to DAP and FOF were 2 and 64, 32 and 128, 1 and 64, and 1 and 32 mg/liter for 8019, 5938, V583, and R7302, respectively. Elevated MICs were observed in 8019 with the DAP 12 and DAP 12 plus FOF simulated regimens (DAP MIC of 16 and 8 mg/liter, respectively). V583 developed nonsusceptibility to DAP (DAP MIC, 16 mg/liter) in DAP 8 and 12 monotherapy regimens. No change in MIC to DAP was observed in R7302. High-level resistance to FOF (MIC of 1,028 mg/liter) developed against all 3 susceptible isolates when utilized as monotherapy.

In vitro PK/PD model.

The average pharmacokinetic parameters observed for FOF were fC_max of 257.9 ± 13.2 mg/liter and half-life of 1.9 ± 0.12 h (targeted, 260 mg/liter and 2.3 h). The average pharmacokinetic parameters observed for DAP 8 were fC_max of 10.22 ± 0.16 mg/liter and half-life of 8.13 ± 0.04 h (targeted, 9.86 mg/liter and 8 h), and for DAP 12 they were fC_max of 14.1 ± 0.12 mg/liter and half-life of 8.98 ± 0.43 h (targeted, 14.7 mg/liter and 8 h).

The in vitro killing activity for simulated antimicrobial regimens is summarized in Table 1. The novel combination of FOF and DAP 12 was significantly more active by 72 h (P ≤ 0.05) than either regimen alone for all isolates examined (Fig. 1). This combination also displayed significantly more activity than monotherapy at 24 h against 5938, V583, and R7302 (P ≤ 0.05). DAP 12 plus FOF demonstrated rapid bactericidal activity sustained to 72 h against 8019, V583, and R7302. Enhancement was observed with the addition of FOF to DAP 12 against all isolates at 24 h and against 8019, V583, and R7302 at 72 h. Improvement was displayed against all isolates with DAP 8 and FOF over either agent alone at 24 h, and this was sustained to 72 h with the exception of 5938. Killing was most impressive against 8019, and the addition of FOF prevented the regrowth seen with DAP 8 and DAP 12 alone. The reduction in log₁₀ CFU/ml demonstrated by DAP 8 in combination with FOF demonstrated an effect similar to that of the combination with DAP 12, but the DAP 8 combination was unable to sustain bactericidal activity against the two E. faecalis strains (V583 and R7302). FOF alone was less active than DAP monotherapy in this model, with the exception of the DAP-nonsusceptible mutant 5938. No regimen achieved bactericidal activity against 5938; however, DAP 12 with FOF resulted in significantly lower colony counts than monotherapy.

TABLE 1.

In vitro activity of DAP alone or in combination with FOF in the one-compartment PK/PD model

Regimena	Activity (log10 CFU/ml) by strain and time pointb
	8019		5938		V583		R7302
	T24	T72	T24	T72	T24	T72	T24	T72
FOF	7.73 ± 0.06	7.81 ± 0.01	6.81 ± 0.39	7.76 ± 0.06	7.95 ± 0.24	7.97 ± 0.57	8.17 ± 0.07	7.60 ± 0.39
D8	4.17 ± 1.28	7.94 ± 0.06	8.08 ± 0.41	8.43 ± 0.30	6.98 ± 0.12	7.75 ± 0.00	7.29 ± 0.04	7.43 ± 0.24
D12	2.36 ± 1.92	7.68 ± 0.21	8.21 ± 0.08	8.54 ± 0.05	7.57 ± 0.01	7.87 ± 0.02	7.11 ± 0.24	7.55 ± 0.51
D8+FOF	2.75 ± 0.64	4.45 ± 0.91*	4.81 ± 0.26*	7.79 ± 0.04	3.51 ± 0.35*	5.88 ± 0.04*	5.13 ± 0.02*	5.69 ± 0.44*
D12+FOF	1 ± 0.00	2.5 ± 1.27*	4.67 ± 0.28*	7 ± 0.17*	3.58 ± 0.06*	3.91 ± 0.21*	5.01 ± 0.57*	4.64 ± 0.02*

^aD8, DAP 8; D12; DAP 12; D8+FOF, DAP 8 with FOF; D12+FOF, DAP 12 with FOF.

^bT24, 24-h time point; T72, 72-h time point. An asterisk indicates significantly greater log₁₀ CFU/ml reduction over monotherapy (P ≤ 0.05).

FIG 1.

Activity of simulated antimicrobial regimens containing daptomycin at 8 mg/kg (DAP-8), daptomycin at 12 mg/kg (DAP-12), fosfomycin (FOF), or growth control (GC) against each isolate in the in vitro one-compartment PK/PD model.

Zeta potential analysis.

Mean cell surface charge of cells in the presence and absence of subinhibitory concentrations of FOF is depicted in Fig. 2. FOF exposure caused significant reductions in surface charge (4.1 to 5.6 mV) for all 4 strains (P < 0.01). Zeta potential of unexposed cells was positively correlated with the DAP MICs (R² = 0.97), indicating that the more susceptible strains had a more negatively charged cell surface than less susceptible ones.

FIG 2.

Zeta potential of tested isolates in the presence or absence of fosfomycin.

Daptomycin-bodipy binding studies.

Daptomycin-bodipy binding studies are illustrated in Fig. 3. Daptomycin-bodipy binding increased by 2.6 times in the presence of fosfomycin and G-6-P (P < 0.0001). There was no change in daptomycin binding observed when cells were exposed to G-6-P.

FIG 3.

Daptomycin-bodipy binding to Enterococcus faecium 8019 alone (A) and in the presence of 25 μg/ml of glucose-6-phosphate (G6P) (B) or G6P and fosfomycin (FOF) (C). Cells were stained with DNA-staining dye DAPI (blue) and bodipy-daptomycin (green). Scale bar, 1 μm.

DISCUSSION

Vancomycin-resistant enterococci cause deep-seated, difficult-to-treat infections that have been associated with increased mortality compared to infections caused by vancomycin-susceptible enterococci (37). These therapeutically challenging infections are frequently complicated by resistance to multiple antimicrobials. Daptomycin is one of the only agents to offer bactericidal activity against VRE, but dose optimization remains a challenge (38). The general consensus among experts is that high doses of daptomycin (greater than 8 mg/kg/day) are necessary to achieve clinical success in the setting of serious enterococcal infections, and emerging data indicate that doses of ≥10 mg/kg/day are necessary for resistance prevention (20, 38–42). Resistance to daptomycin in VRE is an emerging threat that requires immediate attention to derive new therapeutic strategies, such as dose optimization and combination therapy, in order to preserve the clinical utility of this drug.

Our study evaluated the activity of daptomycin in combination with fosfomycin against clinical vancomycin-resistant E. faecium and E. faecalis isolates. The combination of daptomycin and fosfomycin may provide improved killing against daptomycin-nonsusceptible vancomycin-resistant E. faecium and E. faecalis; however, this activity was not sustained in our model. It is of interest that all simulated regimens were less effective against vancomycin-resistant E. faecalis strains than against vancomycin-resistant E. faecium relative to the daptomycin and fosfomycin MIC. This may be related to other adaptations, such as enhanced biofilm production in the E. faecalis strains. Of interest, the fosfomycin-containing regimens prevented the emergence of daptomycin nonsusceptibility in 2/3 daptomycin-susceptible strains, whereas an increase in MIC was seen with daptomycin alone. Similar effects have been seen in methicillin-resistant S. aureus, where the addition of fosfomycin to daptomycin delayed the emergence of resistance to daptomycin in vitro (43). Fosfomycin exposure appeared to reduce cell surface charge, which suggests that the mechanism of synergy is similar to that seen with beta-lactams and daptomycin against enterococci and Staphylococcus aureus (27, 44, 45). Support for this hypothesis comes from recent data, which demonstrate that fosfomycin appears to reduce PBP1 expression in S. aureus (10). PBP1 expression is increased with daptomycin exposure and is believed to be important in the bacterial compensatory response in response to peptide-induced injury (46). Thus, either pharmacologic antagonism of PBP1, as seen with beta-lactam antibiotics, or compromising expression of key compensatory proteins like PBP1 by fosfomycin may be a foundation for increasing daptomycin-mediate bacterial killing efficiency.

Daptomycin susceptibility changes have been found to be associated with changes in cell surface charge in E. faecium and E. faecalis (47, 48). Consistent with a previous study which demonstrated 8019 to be more negatively charged than 5938 through cytochrome c binding and increased susceptibility to human cathelicidin LL37 killing (48), we observed a lower zeta potential in 8019, indicating a more negatively charged surface of 8019 than of 5938. The addition of fosfomycin resulted in a further reduction of cell surface charge, which is a plausible explanation for the enhanced killing demonstrated when combined with daptomycin. VRE resistance to daptomycin has previously been linked to genetic changes in genes that regulate cell envelope homeostasis and membrane lipid metabolism (liaFSR, yycFG, cls, and cfa); however, there appears to be a variety of genetic pathways to daptomycin nonsusceptibility (49–55). In addition to altered cell surface charge, an alternative mechanism that has been proposed is related to redistribution of membrane lipids that drive daptomycin binding away from division septa, where it is believed to be more devastating to the organism (53, 55, 56). Mutations of the phosphotransferase system (PTS), previously associated with bacteriocin resistance in E. faecium and E. faecalis, may also contribute to daptomycin nonsusceptibility in E. faecium (48, 57). This finding is especially interesting in the context of the synergy between fosfomycin and daptomycin since the phosphotransferase system plays a role in G-6-P transport, which is an important cofactor for fosfomycin transport. The nature of the relationship between the mutations in this enzyme system observed in daptomycin nonsusceptible strains and the activity of fosfomycin is unclear but warrants further exploration.

As with other published studies with a similar design, this study has some limitations. Due to the practical limitations of in vitro modeling, we simulated one-compartment pharmacokinetics of the study drugs, which may have slightly overestimated their average clinical exposures. The simplification of multicompartmental clinical pharmacokinetics is common in in vitro modeling, even when the investigators use a two-compartment model, such as a simulated endocardial vegetation model or a hollow-fiber model (58–60). While these strategies may modestly overestimate the drug exposures in the model, due to the short distribution phase of these drugs, which was not simulated, we do not believe this had a significant impact on the results. Other studies have successfully utilized a similar model to describe the pharmacokinetic-pharmacodynamic effect of fosfomycin (61). The primary objective of this study was to evaluate the pharmacodynamic interactions between daptomycin and fosfomycin, and we believe that the subtle differences in actual versus simulated pharmacokinetics should have little impact on the overall pharmacodynamics of this combination or the interpretation of the results of this study.

Conclusions.

Daptomycin plus fosfomycin demonstrated sustained bactericidal activity against daptomycin-susceptible vancomycin-resistant E. faecium in vitro and may provide improved pharmacodynamics against daptomycin-nonsusceptible, vancomycin-resistant E. faecium and E. faecalis. Further research is warranted to evaluate the activity of this combination against other strains of VRE, further elucidate the mechanism(s) of synergy, and determine the potential role for this combination in a clinical setting.