Bacteriophage-Antibiotic Combinations for Enterococcus faecium with Varying Bacteriophage and Daptomycin Susceptibilities

Concerns regarding increased prevalence of daptomycin (DAP)-resistant strains necessitate novel therapies for Enterococcus faecium infections. Obligately lytic bacteriophages are viruses that target, infect, and kill bacterial cells. Limited studies have evaluated phage-antibiotic combinations against E. faecium. After an initial screen of eight E. faecium strains, three strains with varying DAP/phage susceptibilities were selected for further experiments.

ABSTRACT

Concerns regarding increased prevalence of daptomycin (DAP)-resistant strains necessitate novel therapies for Enterococcus faecium infections. Obligately lytic bacteriophages are viruses that target, infect, and kill bacterial cells. Limited studies have evaluated phage-antibiotic combinations against E. faecium. After an initial screen of eight E. faecium strains, three strains with varying DAP/phage susceptibilities were selected for further experiments. Phage-to-strain specificity contributed to synergy with antibiotics by time-kill analyses and was associated with lower development of phage resistance.

INTRODUCTION

Vancomycin-resistant enterococcal (VRE) bloodstream infections have been associated with mortality rates of >30% (1, 2). Enterococcus faecium is the most problematic species given that >80% and >90% of strains have been shown to be vancomycin (VAN)- and ampicillin (AMP)-resistant, respectively (3–6). Vancomycin-resistant E. faecium infections are often treated with daptomycin (DAP), because DAP has potent in vitro bactericidal activity against vancomycin-resistant E. faecium isolates (7). Unfortunately, DAP resistance is reported to emerge during treatment of enterococcal infections, and prevalence of this resistance is increasing (8).

Although mechanisms of DAP resistance have not been fully elucidated, relevant genetic pathways in enterococci have been found to involve mutations in liaFSR (three-component regulatory system involved in cellular membrane stress response) (9–11). Emerging data suggest that E. faecium isolates with mutations that increase proclivity for DAP resistance selection (e.g., liaFSR) may not respond to DAP monotherapy, regardless of DAP exposure (12–14). Antibiotic combination therapy, particularly with β-lactams, has been advocated as an alternative strategy to improve patient outcomes and decrease DAP resistance emergence (15, 16). We have demonstrated that DAP–β-lactam combinations can be highly synergistic against E. faecium infection, but this is strain and β-lactam specific (14; unpublished data). Furthermore, DAP–β-lactam combinations have been shown to fail in clinical practice (17). With limited efficacious options, therapeutic alternatives are urgently needed to treat E. faecium infections refractory to DAP–β-lactam combinations.

Obligately lytic bacteriophages (phages) are viruses that target, infect, and kill bacterial cells (18). Multiple positive interactions have been described with phage-antibiotic combinations, including enhanced bacterial killing and alterations in the emergence of antibiotic/phage resistance (19–21). Although few reports have evaluated Enterococcus faecalis, limited to no studies have evaluated phage-antibiotic combinations against E. faecium, the most difficult-to-treat species of enterococci (22–24).

The objective of this study was to evaluate the ability of phage plus DAP alone and in addition to DAP plus various β-lactams to improve bacterial killing and prevent resistance development in E. faecium strains with varying susceptibilities to DAP and phage. Our primary hypothesis was that initial bacterial phage susceptibility would play a role in synergy with antibiotics and prevention of emergence of phage resistance.

E. faecium bacteriophage 113 (ATCC 19950-B1) and propagating organism E. faecium (ATCC 19950) were purchased commercially from ATCC (Manassas, VA). DAP, AMP, and ertapenem (ERT) powder were purchased commercially from Sigma Chemical Company (St. Louis, MO), and ceftaroline (CPT) analytical powder was obtained from Allergan Pharmaceuticals (Parsippany, NJ). Mueller-Hinton broth II (Difco, Detroit, MI) with 50 mg/liter calcium and 12.5 mg/liter magnesium was used for susceptibility testing and time-kill analyses (TKAs). All MICs were performed following CLSI guidelines (25). The susceptibility of phage ATCC 19950-B1 against eight randomly selected E. faecium strains was evaluated as previously described using a modified small-drop agar overlay method (26). After phage quantification, high, medium, and low phage susceptibility were defined as phage counts of >10⁷, between 10³ and 10⁷, and <10³ PFU/ml, respectively. Phage nonsusceptibility was defined as no visual detection of individual phage plaques and/or no bacterial lawn clearance. Strains fully resistant to the phage were not utilized in further experiments.

TKAs were performed as previously described using inocula that were obtained from stationary-phase cultures and with two replicates obtained at each time point (27). All antimicrobials were tested at 0.5× and 0.25× MIC or maximum concentration of free drug in serum (fC_max), whichever was lower. The fC_max concentrations utilized throughout this evaluation were 17 μg/ml for CPT and 15.5 μg/ml for ERT; AMP fC_max did not need to be utilized. Given that phage-induced bacterial killing is strain specific, phage and antibiotic dose optimizations were performed to evaluate the theoretical multiplicity of infection (MOI) (the number of phages added to the well divided by the number of bacteria within the same well) that produced optimal observations of synergy with antibiotic combinations at the two previously listed suboptimal exposures (0.25× and 0.5× MIC) (28). Any samples that contained phage were centrifuged at 12,000 rpm for 2 min with the supernatant removed and replaced with normal saline (0.9% sodium chloride) to reduce the concentration of unadsorbed phages. Synergy was defined as a ≥2-log₁₀-CFU/ml kill compared to the most effective agent (or double-combination regimen) alone at 24 h. Bactericidal activity was defined as a ≥3-log₁₀-CFU/ml reduction from baseline. The emergence of DAP/phage resistance was determined as previously described by using the 24-h TKA liquid sample (14, 26, 29, 30). Phage resistance was scored from the spots as resistant (no visible effect of phage versus phosphate-buffered saline control), intermediate (bacteria present but phage activity visible), and susceptible (<10 CFU in spot) (30). Phage counts were assessed using a modification of the small-drop agar overlay method (to determine whether the presence [or lack] of antibiotics had an impact of phage growth) (26, 29, 30).

Phage activity is well documented to be bacterial strain specific (28). Noting this, we sought to compare the activity of commonly utilized antibiotics for refractory E. faecium infections in combination with phage for strains that exhibited variability susceptibility to phage and DAP. Strains with low (R496; DAP MIC, 32 μg/ml), medium (HOU503; DAP MIC, 2 μg/ml), and high (R497; DAP MIC, 16 μg/ml) phage susceptibility were selected for additional experiments due to the presence of liaFSR mutations and varying DAP/phage susceptibility. Other antibiotic MICs are listed in Table 1.

TABLE 1.

Initial antimicrobial MICs prior to 24-h time-kill analyses

Antimicrobiala	MIC (μg/ml) for:
	R497 (high susceptibility to phage)	HOU503 (medium susceptibility to phage)	R496 (low susceptibility to phage)
DAP	16	2	32
AMP	128	128	64
CPT	>64	32	>64
ERT	>64	>64	>64

^aDAP, daptomycin; AMP, ampicillin; CPT, ceftaroline; ERT, ertapenem.

Against DAP-resistant R497 harboring liaFSR substitutions, synergistic/bactericidal effects were seen with DAP-AMP-phage and DAP-CPT-phage, whereas DAP-phage exhibited synergy. Synergistic effects against HOU503 (a DAP-tolerant isolate, also with mutations in liaFSR) were noted with DAP-CPT-phage and DAP-ERT-phage. No enhancement was noted with phage addition to antibiotics against R496 (DAP-resistant isolate with mutations in liaFSR) (Fig. 1). No further DAP resistance emerged in R496 or R497 in any TKA regimens. However, DAP MICs of HOU503 increased from 2 to 32 μg/ml for all regimens except DAP-AMP-phage. We have previously shown in a pharmacokinetic/pharmacodynamic simulated endocardial vegetation (SEV) model that AMP, CPT, and ERT prevented the emergence of DAP resistance in HOU503 at humanized dose exposures (14). We performed TKA at the same humanized exposures, and the emergence of DAP resistance was prevented when β-lactam combination therapy was utilized. The development of phage resistance was less pronounced in strains with higher initial phage susceptibility (Table 2). No meaningful differences were observed in phage quantification. (Supplemental tables may be found at https://www.dropbox.com/s/8ah3ftnocbervb7/E.%20fcm%20-%20Phage%20-%20Supplemental.pdf?dl=0.)

FIG 1.

Time-kill analyses of DAP-AMP, DAP-CPT, and DAP-ERT alone and in combination with bacteriophage against R497 at 0.5× MIC or fC_max (DAP, 8 μg/ml; AMP, 64 μg/ml; CPT, 17 μg/ml; ERT, 15.5 μg/ml) and a theoretical MOI of 0.1, HOU503 at 0.25× MIC or fC_max (DAP, 0.5 μg/ml; AMP, 32 μg/ml; CPT, 8 μg/ml; ERT, 15.5 μg/ml) and a theoretical MOI of 1.0, and R496 at 0.25× MIC or fC_max (DAP, 8 μg/ml; AMP, 16 μg/ml; CPT,17 μg/ml; ERT, 15.5 μg/ml) and a theoretical MOI of 1.0. Synergy was defined as a ≥2-log₁₀-CFU/ml kill compared to the most effective agent (or double-combination regimen) alone at 24 h. Bactericidal activity was defined as a ≥3-log₁₀-CFU/ml reduction from baseline. For R497, bactericidal and synergistic activity was noted with the addition of phage to DAP-AMP (A) and DAP-CPT (B), whereas synergistic activity was noted with the addition of phage to DAP (A to C). For HOU503, synergistic effects were noted with the triple combinations DAP-CPT-phage (B) and DAP-ERT-phage (C). For R496, no enhancement was noted with the addition of phage to DAP-AMP (A), DAP-CPT (B), or DAP-ERT (C). Values are means ± standard deviations.

TABLE 2.

Phage susceptibility after 24-h time-kill analyses

Regimena	Phage resistanceb for:
	R497	HOU503	R496
P	I	I	R
P-DAP	I	I	R
P-AMP	I	I	R
P-CPT	I	I	R
P-ERT	I	R	R
P-DAP-AMP	S	I	R
P-DAP-CPT	S	R	R
P-DAP-ERT	S	R	R

^aDAP, daptomycin; AMP, ampicillin; CPT, ceftaroline; ERT, ertapenem; P, phage.

^bS, susceptible; I, intermediate; R, resistant.

Enterococcus spp. isolates are leading causes of nosocomial infections internationally, with multidrug-resistant (MDR) strains increasing in prevalence. Daptomycin is a key therapeutic option against these infections. However, the increase in MDR E. faecium strains with reduced susceptibility to VAN, AMP, and DAP coupled with a sparse antibiotic development pipeline for VRE has led to many challenging clinical scenarios. Combination therapy with β-lactams has been advocated as an alternative strategy (16). Although positive interactions have been described, these enhancements have been shown to be strain and β-lactam specific (14, 31–33; unpublished data). This is extremely problematic in the clinical setting, because most practitioners utilize a “best guess” scenario when choosing which β-lactam to add upfront or in recalcitrant infections.

In light of some failures observed with antibiotic combination therapy, the novel addition of phages to antibiotics has been promoted as another alternative strategy. Our report adds to the growing body of literature describing positive interactions with phage-antibiotic combinations (18–21). Not only did we show synergistic activity with phage in addition to two antibiotics, but we also showed synergy with DAP-phage against a DAP-resistant strain (R497) and that phage-to-strain specificity contributed to synergy with antibiotics. Although we did not show that subinhibitory concentrations of antibiotics fully prevented the emergence of phage resistance in all strains tested, there seemed to be a relationship between the initial phage susceptibility of each strain and the subsequent emergence of phage resistance.

Limitations of this study include utilizing monophage therapy in comparison to evaluating phage cocktails. Furthermore, although we performed phage susceptibility testing on eight E. faecium strains, we only conducted further examinations on three strains. Finally, it is important to note that traditional methodology of TKA entails dosing antimicrobials at the beginning of the experiments without repeated administration. Whether repeated phage administrations would enhance bacterial killing or prevent the emergence of resistance in these experiments is yet to be elucidated.

In summary, we analyzed phage specificity as it relates to synergistic activity with antibiotics against varying resistant strains of E. faecium as proof of concept. A phage was used that showed variable susceptibility to the strains of interest, which allowed for the observation of differences in efficacy of bacterial killing and resistance responses. The findings of this study further support that phages with enhanced activity to bacterial strains of interest may be useful in combination with antibiotics in the clinical setting. Further in vitro examinations should be performed to fully understand phage-antibiotic interactions and how these results will translate in vivo, in addition to working to develop tools to optimize the selection of antimicrobial combinations to use upfront in high-risk patients or in recalcitrant infections.