Characterization of High-Level Daptomycin Resistance in Viridans Group Streptococci Developed upon In Vitro Exposure to Daptomycin

Abstract

Viridans group streptococci (VGS) are part of the normal flora that may cause bacteremia, often leading to endocarditis. We evaluated daptomycin against four clinical strains of VGS (MICs = 1 or 2 μg/ml) using an in vitro-simulated endocardial vegetation model, a simulated bacteremia model, and kill curves. Daptomycin exposure was simulated at 6 mg/kg of body weight and 8 mg/kg every 24 h for endocardial and bacteremia models. Total drug concentrations were used for analyses containing protein (albumin and pooled human serum), and free (unbound) drug concentrations (93% protein bound) were used for analyses not containing protein. Daptomycin MICs in the presence of protein were significantly higher than those in the absence of protein. Despite MICs below or at the susceptible breakpoint, all daptomycin regimens demonstrated limited kill in both pharmacodynamic models. A reduction of approximately 1 to 2 log₁₀ CFU was seen for all isolates and dosages except daptomycin at 6 mg/kg, which achieved a reduction of 2.7 log₁₀ CFU/g against one strain (Streptococcus gordonii 1649) in the endocardial model. Activity was similar in both pharmacodynamic models in the presence or absence of protein. Similar activity was noted in the kill curves over all multiples of the MIC. Regrowth by 24 h was seen even at 8× MIC. Postexposure daptomycin MICs for both pharmacodynamic models increased to >256 μg/ml for all isolates by 24 and 72 h. Despite susceptibility to daptomycin by standard MIC methods, these VGS developed high-level daptomycin resistance (HLDR) after a short duration following drug exposure not attributed to modification or inactivation of daptomycin. Further evaluation is warranted to determine the mechanism of resistance and clinical implications.

INTRODUCTION

Viridans group streptococci (VGS) include a number of species and are commensal Gram-positive bacteria often responsible for human disease, particularly infective endocarditis and bacteremia in neutropenic patients (1–5). Historically, these organisms have been relatively susceptible to most antibiotics. However, the frequency of multidrug resistance has been increasing (6–17). Much of the resistance and virulence associated with these organisms can be exchanged among multiple streptococcal species via horizontal transfer of genetic material (13, 18, 19). Antibiotic resistance has been reported primarily for penicillin, macrolides, and fluoroquinolones (4, 6, 7, 9, 12). The resistance related to the mef and erm genes, which confer resistance to macrolides, lincosamides, and streptogramin B, has been well characterized in VGS (3, 4, 7, 12). More recently, reports of daptomycin and linezolid resistance have also been published; these reports include both in vitro studies and in vivo studies with 2 clinical cases for each agent (20–25). Vancomycin resistance in VGS has not previously been identified (10, 11, 25); however, recently a Streptococcus anginosus isolate was found to contain the vanG element, conferring resistance to vancomycin in this strain (55).

Daptomycin (DAP) is a cyclic lipopeptide antibiotic with activity against multidrug-resistant Gram-positive organisms, including Staphylococcus aureus resistant to vancomycin (VAN) and linezolid (LZD)- and vancomycin-resistant enterococci. To date, DAP has not exhibited any cross-resistance with other antibiotic classes nor has any plasmid-mediated resistance occurred (20). As Gram-positive bacterial resistance increases, empirical usage of DAP continues to rise, particularly in patients at high risk of endocarditis or S. aureus infection. With VGS as a common cause of endocarditis and limited data with DAP against these strains, we evaluated the in vitro activity of DAP at various dosages (6 and 8 mg/kg of body weight) or multiples of the MIC against four clinical isolates of VGS with elevated daptomycin MICs (1 or 2 μg/ml).

(Portions of this study were presented at the 21st European Congress of Clinical Microbiology and Infectious Diseases [ECCMID], May 2011, Milan, Italy.)

MATERIALS AND METHODS

Bacterial strains.

Four clinical strains of VGS, all in the mitis subgroup (Streptococcus mitis strain 1643, Streptococcus oralis strains 1647 and 1648, and Streptococcus gordonii strain 1649) were evaluated. These strains, provided by Cubist Pharmaceuticals (Lexington, MA), obtained from Sentry Surveillance archived cultures (JMI Laboratories, North Liberty, IA) collected between 1999 and 2003, were isolated from hospitalized patients with infective endocarditis. The isolates were selected based on a MIC to DAP at the high end of the wild-type distribution (up to 2 μg/ml). Isolates were verified by Molecular Epidemiology Inc. (MEI) (Lake Forest Park, WA) using 16S rRNA analysis supported by multiple phenotypic tests, including but not limited to Gram stain, catalase/oxidase reaction, multiple sugar fermentation, and optochin/bacitracin susceptibility. The reference strains for pre- and postexposure isolates are listed in Table 1. Preexposure isolates refer to all testing or results prior to exposure to daptomycin, and postexposure isolates refer to all testing or results subsequent to any treatment-based exposure to daptomycin.

TABLE 1.

Summary of pre- and postexposure MIC testing results under standard testing conditions

Reference strain	Genetic IDa	Pre/postexposure statusb	MIC (μg/ml)
			DAPb,c	DAPd,e	RIFb,c	VANd	GENc,e	LZDf
1643	S. mitis	Pre	1	NA	0.5	0.5	4	1
5064	S. mitis	Post	>32	>256	≤1	0.5	NA	1
1647	S. oralis	Pre	2	NA	0.5	0.5	4	1
5065	S. oralis	Post	>32	>256	≤1	0.5	NA	1
1648	S. oralis	Pre	1	NA	256	0.5	4	1
5066	S. oralis	Post	>32	>256	>4	0.5	NA	1
1649	S. gordonii	Pre	2	NA	0.125	0.5	1	1
5067	S. gordonii	Post	>32	>256	>4	0.5	NA	1

^aGenetically identified species.

^bPostexposure strains were collected after 72 h of daptomycin treatment at 8 mg/kg/day.

^cMICs were determined by the investigator using broth microdilution (preexposure strains) or by Laboratory Specialists, Inc. (Westlake, OH) (postexposure strains).

^dMICs were determined by the investigator using Etest methodology (postexposure strains).

^eNA, not available (MICs were not determined for this organism and antibiotic).

^fAll MICs were determined by Laboratory Specialists, Inc., using broth microdilution.

Antibiotics.

DAP (lot no. 850753A, CDCX01, MCB2009) analytical powder was provided by Cubist Pharmaceuticals (Lexington, MA). LZD was purchased commercially from the manufacturer. VAN, rifampin (RIF), and gentamicin (GEN) were purchased from Sigma-Aldrich (St. Louis, MO).

Media.

Multiple medium broths were studied to determine the most appropriate media to conduct time-kill analysis with these strains based on the strict medium requirements for VGS growth and calcium content for DAP activity. Four types of broth (Todd-Hewitt broth with and without 5% yeast extract, Columbia broth, brain heart infusion broth, and Mueller-Hinton broth) were tested for their free (unbound) calcium concentration with and without 50 mg/liter calcium supplementation (Laboratory Specialists, Inc., Westlake, Ohio) via a calcium probe. The concentrations of free calcium for each broth type are listed in Table 2 with MICs determined for DAP against all strains and VAN for a representative strain to determine the impact of broth type on antibiotic susceptibility. There was a significant increase in DAP MICs noted in all broth types (except calcium-supplemented Mueller-Hinton broth) even with sufficient calcium supplementation up to 50 mg/liter. Based on this analysis and DAP dependency on calcium for activity, all DAP models utilized calcium-supplemented Mueller-Hinton broth (CA-SMHB) with 50 mg/liter calcium and 12.5 mg/liter magnesium (26–29). All other experiments utilized Mueller-Hinton broth (Difco, Detroit, Michigan) supplemented with 25 mg/liter calcium and 12.5 mg/liter magnesium (SMHB). Colony counts were determined using tryptic soy agar with 5% sheep blood (TSA-SB) (Becton, Dickinson and Company, Sparks, MD) plates.

TABLE 2.

Free calcium concentration with MIC comparisons of VGS in different media

Mediuma	Free Ca+ concn	MIC (μg/ml) for the following drug/reference strainb:
		DAP/1643	DAP/1647	DAP/1648	DAP/1649	VAN/1648
CA-SMHB	56.608	1	2	1	2	0.5
THB	5.370	64	>64	64	64	2
STHB	16.565	4	>64	64	8	1
THB+Y	29.079	32	>64	>64	64	1
STHB+Y	17.133	4	>64	16	8	2
CB	7.055	64	>64	>64	>64	0.5
SCB	16.565	8	64	16	16	1
BHI	4.669	>64	>64	>64	>64	1
SBHI	35.252	4	>64	32	16	1

^aThe different media tested were as follows: CA-SMHB, calcium-supplemented Mueller-Hinton broth with 50 mg/liter Ca and 12.5 mg/liter Mg; THB, Todd-Hewitt broth; STHB, Todd-Hewitt broth with 50 mg/liter Ca and 12.5 mg/liter Mg; THB+Y, Todd-Hewitt broth plus 5% yeast extract; STHB+Y, Todd-Hewitt broth with 50 mg/liter Ca and 12.5 mg/liter Mg plus 5% yeast extract; CB, Columbia broth; SCB, Columbia broth with 50 mg/liter Ca and 12.5 mg/liter Mg; BHI, brain heart infusion broth; SBHI, brain heart infusion broth with 50 mg/liter Ca and 12.5 mg/liter Mg.

^bThe drugs were daptomycin (DAP) and vancomycin (VAN). The organisms tested were S. mitis strain 1643, S. oralis strains 1647 and 1648, and S. gordonii strain 1649.

Susceptibility testing.

MICs and minimum bactericidal concentrations (MBCs) of the antibiotics were determined by broth microdilution in SMHB (28). DAP MICs and MBCs were determined in supplemented broth (CA-SMHB) as described above. DAP MICs and MBCs were also determined in the presence of albumin (3.5 to 4 g/dl) (human albumin 25%; CSL Behring LLC, Kankakee, IL) and pooled human serum (PHS). Five-microliter samples from clear wells in the MIC experiments were plated onto TSA-SB plates for determination of MBCs, and all samples were incubated at 35°C for 24 h. Postexposure isolate MICs were also determined by Etest or broth microdilution methodology.

SEVs.

Organism stocks were prepared by inoculating 5-ml test tubes of SMHB with colonies harvested from fresh overnight growth on TSA-SB. The test tubes were then incubated for 24 h on a rotator at 35°C. The test tubes were then centrifuged for 15 min at 3,500 × g, and the supernatant was removed. The remaining pellet of organism was collected and resuspended to achieve a concentration of 1 × 10¹⁰ CFU/ml. Simulated endocardial vegetations (SEVs) were prepared by mixing 0.1 ml of organism suspension (final inoculum, 1 × 10⁹ CFU/0.5 g), 0.4 ml of human cryoprecipitate from volunteer donors (Gulf Coast Regional Blood Center, Houston, TX), and 0.025 ml of platelet suspension (platelets mixed with normal saline, 250,000 to 500,000 platelets per clot in 1.5-ml siliconized Eppendorf tubes; platelets obtained and pooled from volunteer donors, LifeShare Blood Center, Shreveport, LA). Bovine thrombin (5,000 units/ml; lot R114A1048, Jones Pharma Inc., St. Louis, MO) (0.05 ml) was added to each tube, after insertion of a sterile monofilament line into the mixture. The resultant simulated vegetation was then removed from the Eppendorf tubes with a sterile 21-gauge needle.

Simulated endocardial vegetation model (SEVM).

An in vitro infection model consisting of a 250-ml one-compartment glass apparatus with sample ports incorporated, in which the simulated fibrin clots were suspended and sealed with a rubber stopper, was utilized. The apparatus was prefilled with SMHB or CA-SMHB, and antibiotics were administered as boluses over a 72-h period into the central compartment via an injection port. The model apparatus was placed in a 35°C water bath throughout the procedure, and a magnetic stir bar was placed in the media 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) set to simulate the half-lives of the antibiotics. The pH was monitored throughout all experiments with DAP due to the possible effects on its activity (26). SEVs were removed at various time points over a 72-h period. Samples were homogenized, diluted, plated onto TSA-SB, and incubated at 35°C for 24 h. Colonies were counted after 24 h of incubation, and the number of CFU/g was calculated. DAP was administered at dosages of 6 mg/kg or 8 mg/kg every 24 h for an estimated human total peak concentration/trough concentration of 80/10 and 133/17 μg/ml, respectively, with an average half-life of 8 h and 93% protein binding. Total drug concentrations were simulated, due to the significant effects of protein binding on DAP and since each SEV contains an average total protein concentration of 6.8 to 7.4 g/dl and 3.0 to 3.5 g/dl of albumin (26, 30). LZD was given in 600-mg doses every 12 h with an estimated peak concentration/trough concentration of 18.3/3.6 μg/ml and an average half-life of 5 h and 31% protein binding (31). VAN was administered to simulate 1 g every 12 h for peak concentration/trough concentration of 30 to 40/10 to 15 μg/ml with a half-life of 6 h and 55% protein binding.

All infection model experiments were performed in triplicate to ensure reproducibility. In addition, model experiments in the absence of antibiotics were performed over 72 h to ensure adequate growth of the organisms in the model.

SBM.

An in vitro pharmacodynamic infection model consisting of a one-compartment glass chamber with multiple ports for the removal of medium, delivery of antibiotics, and collection of bacterial and antimicrobial samples was utilized to determine the in vitro activity of the antibiotic regimens without any protein components. The apparatus was prefilled with SMHB or CA-SMHB, and antibiotics were administered as boluses over a 72-h period. The simulated bacteremia model (SBM) was placed in a 35°C water bath for the duration of the experiment, and a magnetic stir bar was placed in the medium for continuous mixing of the medium. A peristaltic pump (Masterflex; Cole-Parmer Instrument Company, Chicago, IL) was used to replace antibiotic-containing medium with fresh medium at a rate to clear the various antibiotics in patients with normal renal function by utilizing the drugs' half-lives. The pH was monitored throughout all experiments with DAP due to the possible effects on its activity (26). Bacterial colonies after overnight growth on TSA-SB were added to SMHB to obtain a 0.5 McFarland suspension. An aliquot of this suspension was added to the model to produce an initial inoculum of approximately 1 × 10⁶ CFU/ml. Colony counts were measured at 0 (baseline), 8, 24, 32, 48, and 72 h after introduction of antibiotic. Bacterial density in CFU/ml was calculated after 24 h of incubation at 35°C. Free (unbound) drug concentrations were simulated in the SBM, since there is no protein within the model based on the dosages, half-lives, and protein binding indicated above in the SEVM. Free drug concentrations used were as follows. The DAP fC_max/fC_min values (maximum concentration of the free, unbound fraction of a drug/minimum concentration of the free, unbound fraction of a drug) were 5.6/0.7 and 9.3/1.2 μg/ml for a 6-mg/kg and 8-mg/kg dose given every 24 h, respectively. The LZD fC_max/fC_min was 12.6/2.5 for a 600-mg dose every 12 h, and the VAN fC_max/fC_min was 15.8/5.6 for a 1-g dose every 12 h. All infection model experiments were performed in triplicate to ensure reproducibility. In addition, model experiments with no antibiotics were performed over 72 h to ensure adequate growth of the organisms in these models.

FIC synergy determination.

Approximately 1 × 10⁷ CFU/ml of each organism was pipetted into microtiter assay plates containing concentrations of each antibiotic combination (DAP plus GEN and DAP plus RIF) ranging from 1/32 MIC to 4× MIC. Microtiter plates were incubated at 35°C for 24 h. After 6, 12, and 24 h, the plates were read for detection of inhibition of bacterial growth. The experiments were performed in triplicate. The fractional inhibitory concentration (FIC) indices were calculated by the following formulas: FIC of drug A = MIC of drug A in combination/MIC of drug A alone, FIC of drug B = MIC of drug B in combination/MIC of drug B alone, and FIC index = FIC of drug A + FIC of drug B. Synergy was defined as an FIC index of ≤0.5, an additive effect or indifference was defined as an FIC index of >0.5 to ≤4, and antagonism was defined as an FIC index of >4 (32). Synergy screening was conducted on each antibiotic pair.

Pharmacokinetic analysis.

Samples were obtained from broth (0.5 ml from each infection model), through the injection port, at 0.5, 1, 2, 4, 8, 24, 32, 48, and 72 h for determination of the antibiotic concentrations based on a pharmacodynamic model (SEVM or SBM). Samples were also taken from the homogenized SEVs obtained at 0, 8, 24, 32, 48, and 72 h for determination of the antibiotic concentration within the vegetation. Samples were stored at −70°C until analysis. DAP concentrations were determined by microbioassay utilizing Micrococcus luteus ATCC 9341 as described below (33). The vegetation drug concentrations were also determined via microbioassay using the same indicator organism. Blank 0.25-in. disks were spotted with 20 μl of the standards or samples. Each standard was tested in triplicate by placing the disk on Antibiotic Assay Medium 1 agar plates, which was preswabbed with a 0.5 McFarland suspension of the test organism. The plates were incubated for 18 to 24 h at 35°C, at which time the zone sizes were measured. The standard DAP antibiotic concentrations used were 150, 100, and 10 μg/ml for total drug concentration dosing and 10, 2.5, and 1.25 μg/ml for free (unbound) drug concentration dosing, with the lower limit of detection of 1.25 μg/ml due to the limitation of the blank disk size. The intraday and interday coefficients of variation percentage (CV%) for the DAP microbioassay were ≤7.4 and ≤9.8 for total drug concentration standards and ≤9.1 and ≤10.4 for free drug concentration standards, respectively. VAN concentrations were determined by fluorescence polarization immunoassay (TDx assay; Abbott Diagnostics, Irving, TX) with a lower limit of detection of 2.0 μg/ml and an intraday CV% of ≤5.4. LZD concentrations were determined by high-performance liquid chromatography (HPLC) as previously described (34) with a lower level of detection of 0.5 μg/ml and an intraday and interday CV% of ≤5.7 and ≤11.8. The antibiotic peak and trough concentrations, half-lives, and area under the curve were determined by the trapezoidal method using PK Analyst (version 1.10; MicroMath Scientific Software, Salt Lake City, UT).

Pharmacodynamic analysis.

Three vegetations were removed from each model (for a total of nine vegetations) at 0, 8, 24, 32, 48, and 72 h. Vegetations were homogenized and diluted in cold saline, and 20 μl of the homogenate was plated in triplicate onto TSA-SB plates. The plates were incubated at 35°C for 24 h, at which time colony counts were performed. The total reduction in log₁₀ CFU/g over 72 h was determined by plotting time-kill curves based on the number of remaining organisms over the 72-h time period. The time to achieve a 99.9% bacterial load reduction was determined by linear regression (if r² ≥ 0.9) or by visual inspection. The C_max-to-MIC ratio (C_max/MIC), time above the MIC for 24 h (T > MIC₂₄), and the area under the concentration-time curve from 0 to 24 h (AUC_0–24)/MIC ratio were determined for the different dosing regimens.

Resistance.

The 100-μl samples from each time point were plated on TSA-SB plates containing 4× and 8× the MIC of the respective antibiotic to assess the development of resistance. The plates were then examined for growth after 48 h of incubation at 35°C. Development of resistance was evaluated for each model at the 24-, 48-, and 72-h time points. Any growth of organisms observed on the antibiotic-containing plates after 48 h of incubation was considered to show resistance. If resistance developed, further MIC and MBC testing was performed to determine the level of resistance. Pre- and postexposure strains were also subtyped and analyzed for clonal relatedness by restriction fragment length polymorphism (RFLP) using either XbaI or ApaI restriction endonuclease followed by pulsed-field gel electrophoresis (PFGE) at MEI.

Statistical analysis.

Changes in the CFU/g for SEVM or CFU/ml for SBM at 72 h were compared by two-way analysis of variance with Tukey's post hoc test. A P value of ≤0.05 was considered significant. All statistical analyses were performed using SPSS statistical software (release 21.0; SPSS, Inc., Chicago, IL).

Time-kill curve experiments.

Upon completion of SEVMs and SBMs, time-kill curve experiments (KCs) were utilized with DAP only to allow for further evaluation of the extent of resistance development with exposure to multiple drug concentrations (up to 8× MIC) and to examine whether differences existed between studies containing protein (albumin and pooled human serum) and studies not containing protein. KCs were performed by mixing colonies from an overnight growth of organism on TSA-SB plates into an appropriate volume of CA-SMHB to obtain a 0.5 McFarland suspension. This suspension was then diluted to achieve a starting inoculum of 1 × 10⁶ CFU/ml. The following conditions were compared: broth alone, broth plus albumin (3.5 to 4 g/dl), and broth plus pooled human serum (PHS) (50:50). The DAP concentrations run in the KCs were therapeutic (6 mg/kg/day) and multiples of 0.5×, 1×, 2×, 4×, and 8× MIC. Total DAP concentrations (26, 30) were used in all protein-containing experiments, and free drug concentrations were used in experiments containing broth alone. Each regimen was tested in triplicate. Samples (100 μl) were taken at 0, 4, 8, and 24 h, serially diluted with cold normal saline, and plated in triplicate on drug-free TSA-SB plates and drug-containing (2×, 4×, and 8× DAP MICs) TSA with 5% lysed horse blood plates for determination of the numbers of CFU/ml. Colony counts on drug-free and drug-containing plates were compared over the experimental period. For situations in which use of the first dilution was necessary for bacterial enumeration, samples were filtered through a 0.45-μm-pore-size polysulfone filter and washed with cold normal saline, and the filter was applied aseptically to a TSA-SB plate to minimize the potential effects of antibiotic carryover.

Resistance characterization.

Numerous studies were performed on postexposure resistant isolates to characterize the mechanism of resistance.

(i) Determination of heterogeneous resistance in preexposure strains.

Preexposure strains were screened for heterogeneous DAP resistance via a modified macro-Etest and by direct plating of high inocula on daptomycin-containing agar media.

A modified DAP macro-Etest was performed on the preexposure strains according to methods previously described for the detection of heterogeneous vancomycin-intermediate Staphylococcus aureus (hVISA) (35–37). The modified method substituted Mueller-Hinton agar supplemented with 5% sheep blood (MHAB) to support the growth of VGS in place of brain heart infusion agar. Briefly, 200 μl of a 2 McFarland inoculum was placed on MHAB, streaked in three directions with a sterile swab, and allowed to dry for at least 5 to 10 min. A DAP Etest was then added to the plate and incubated for 24 to 48 h at 35°C and 5% CO₂. At both 24 and 48 h, the plates were examined for colonies within the ellipse (zone of inhibition). The presence of colonies within the ellipse indicated a positive screen, while no colonies indicated a negative screen (Fig. 1).

FIG 1.

Modified DAP macro-Etest for S. oralis 1648 with a subsequent standard Etest on the non-DAP-susceptible subpopulation. (a) Modified macro-Etest demonstrating non-DAP-susceptible subpopulations in S. oralis 1648 at 48 h. (b) Standard Etest MIC of a purified colony (circled in panel a and labeled with a dagger) from modified DAP macro-Etest of S. oralis 1648 at 48 h (DAP MIC > 256 μg/ml).

Heterogeneity screens for other antibiotics used in the in vitro pharmacodynamic models and synergy assays (i.e., LZD, VAN, and RIF) were performed for the S. oralis 1647 and 1648 strains using the same methodology.

(ii) Heterogeneous resistance.

Heterogeneous DAP resistance was also examined by plating a high inoculum on DAP-containing MHAB and incubating the culture overnight at 35°C and 5% CO₂. Briefly, an overnight culture of each organism was suspended in 1× phosphate-buffered saline (PBS) and serially diluted. A volume of 100 μl of the undiluted cell suspension and 100 μl of the first 10-fold dilution was spread on each of three plates at each DAP concentration, 0, 4, 8, 16, and 64 μg/ml of daptomycin, respectively. The plates were incubated at 35°C and 5% CO₂ and examined at 24 and 48 h. Growth was subcultured on drug-free MHAB from each plate (15 colonies per DAP concentration). MIC values were then determined by standard Etest methodology in accordance with the manufacturers' recommendations.

(iii) Resistance stability.

The stability of resistance in postexposure strains was assessed via serial passage. Strains from frozen glycerol stocks were cultured onto drug-free MHAB plates and incubated for 24 h at 35°C and 5% CO₂. The following day, the cells were resuspended in 1× PBS, and 100 μl of the cell preparation was inoculated into 10 ml of medium (CA-SMHB) containing 0, 4, or 32 μg/ml of DAP. Cultures were incubated for 24 h at 35°C and 5% CO₂ and were considered the first passage.

(iv) Cross-resistance.

The development of cross-resistance to other antibiotics was examined in S. oralis strain 1647 and its postexposure derivative (S. oralis 5065). MIC testing was performed by broth microdilution in accordance with CLSI guidelines for other cell wall active compounds, such as valinomycin, vancomycin, oxacillin, nisin, bacitracin, and gramicidin.

(v) Mechanism of resistance.

The ability of the VGS strains to inactivate or modify DAP was investigated through a microbioassay and by HPLC. S. oralis 5065, the postexposure derivative of S. oralis 1647, was grown overnight on MHAB containing 10 μg/ml of DAP at 35°C and 5% CO₂. The following day, growth was diluted in 1× PBS to an optical density at 625 nm (OD₆₂₅) of 0.1, and 30 μl of the cell preparation was then inoculated into 3-ml volumes of CA-SMHB containing DAP concentrations of 0, 32, 64 and 128 μg/ml. Two hundred microliters of culture supernatant for preexposure (0 h) and 6 and 24 h after daptomycin exposure was spot plated on a lawn of Streptococcus pneumoniae ATCC 49619 grown to confluence on MHAB. Microbioassay plates were incubated at 35°C and 5% CO₂ and examined 24 h later for zones of inhibition indicating DAP activity. One hundred microliters of the same supernatant was also subjected to HPLC analysis in an Agilent 1100 series Chemstation HPLC. HPLC traces were not subsequently examined to determine the presence of breakdown product.

RESULTS

Susceptibility testing.

MICs were determined for all pre- and postexposure isolates under standard testing conditions, using either broth microdilution or Etest (Table 1). The MICs/MBCs (μg/ml) of DAP tested in CA-SMHB were 1/16, 2/8, 1/8, and 2/16 for strain 1643, 1647, 1648, and 1649, respectively, with tolerance noted in strain 1648 with a ratio of 16. When the antibiotics were tested in the presence of albumin or PHS, the MICs increased to 256 μg/ml for all organisms except for strain 1648, which had an MIC of 512 μg/ml. The MBCs were slightly lower in the presence of albumin, with an MBC of 512 μg/ml for all strains, except for strain 1643, for which it was 1,024 μg/ml. In the presence of PHS, the MBC was 2,048 μg/ml for all organisms except strain 1643 (1,024 μg/ml). The MIC/MBC ratio was elevated for all organisms, ranging from 4 to 16 when the drugs were tested in the absence of protein and 2 to 8 when the drugs were tested in the presence of protein.

Pharmacokinetics.

The pharmacokinetic parameters achieved for DAP, VAN, and LZD are listed in Table 3.

TABLE 3.

Pharmacokinetic parameters for SEVM and SBM

Parameter and regimen	Mean Cmaxa (mg/liter) (SD)	Mean Cminb (mg/liter) (SD)	Mean half-life (h) (SD)
Total concns (SEVM)
DAP (6 mg/kg)	83.2 (6.3)	11.3 (3.7)	8.6 (10.2)
DAP (8 mg/kg)	138.4 (9.8)	18.6 (5.4)	7.5 (0.9)
VAN	38.1 (2.1)	11.5 (1.3)	7.4 (2.3)
LZD	19.6 (4.3)	3.2 (1.4)	4.6 (1.1)
Free concns (SBM)
DAP (6 mg/kg)	6.03 (1.05)	0.96 (0.5)c	8.9 (1.5)
DAP (8 mg/kg)	9.0 (1.5)	1.4 (0.6)c	7.7 (1.8)
VAN	16.6 (1.9)	5.9 (1.5)	7.0 (1.4)
LZD	13.3 (2.7)	2.03 (0.9)	5.2 (0.7)

^aC_max, maximum concentration.

^bC_min, minimum concentration.

^cExtrapolated values.

Pharmacodynamics.

The change in the initial and final bacterial inocula for the SEVM is shown in Table 4. A negative value shows that bacteria were killed, and a positive value reflects bacterial growth. The majority of DAP regimens achieved limited kill (0 to 2 log₁₀ CFU/g reduction) with the greatest activity seen with 6 mg/kg DAP against strain 1649 (2.7 log₁₀ CFU/g reduction). However, two regimens and strains (6 mg/kg DAP against strain 1643 and 8 mg/kg DAP against strain 1648) had an increase in the final colony count for 72 h compared to 0 h. VAN demonstrated more activity and consistency in kill than DAP or LZD with ≥3 log₁₀ CFU/g reduction for two strains (1643 and 1648) and 1.6 and 2.3 log₁₀ CFU/g reduction for strains 1647 and 1649, respectively. LZD was static for all strains with a 1 to 2 log₁₀ CFU/g reduction. A summary of activity in the SEVM is displayed in Fig. 2.

TABLE 4.

Comparison of pharmacodynamic changes for in vitro models (SEVM and SBM)

Model and reference	Regimen	Initial inoculuma	Final inoculuma	Change in log10 CFU/g (SEVM) or log10 CFU/ml (SBM)
SEVM
S. mitis 1643	DAP (6 mg/kg)	8.22 ± 0.11	8.72 ± 0.24	+0.5
	DAP (8 mg/kg)	8.0 ± 0.34	7.93 ± 0.60	−0.07
	VAN	8.2 ± 0.16	4.81 ± 0.76	−3.39
	LZD	7.6 ± 0.16	6.13 ± 0.12	−1.47
	None (growth control)	8.0 ± 0.22	8.4 ± 0.4	+0.4
S. oralis 1647	DAP (6 mg/kg)	8.51 ± 0.13	8.12 ± 0.21	−0.39
	DAP (8 mg/kg)	9.02 ± 0.03	7.32 ± 0.32	−1.7
	VAN	8.13 ± 0.32	6.49 ± 0.73	−1.64
	LZD	7.5 ± 0.7	6.12 ± 0.31	−1.38
	None (growth control)	8.43 ± 0.13	8.2 ± 0.4	−0.23
S. oralis 1648	DAP (6 mg/kg)	8.96 ± 0.11	8.1 ± 0.07	−0.86
	DAP (8 mg/kg)	8.41 ± 0.13	8.64 ± 0.14	+0.23
	VAN	8.6 ± 0.12	5.54 ± 0.34	−3.06
	LZD	7.97 ± 0.13	5.85 ± 0.84	−2.12
	None (growth control)	8.3 ± 0.06	8.18 ± 0.3	−0.12
S. gordonii 1649	DAP (6 mg/kg)	8.7 ± 0.51	6.02 ± 0.23	−2.68
	DAP (8 mg/kg)	8.8 ± 0.06	7.65 ± 0.24	−1.15
	VAN	8.78 ± 0.08	6.51 ± 0.15	−2.27
	LZD	8.56 ± 0.05	6.34 ± 0.44	−2.22
	None (growth control)	7.6 ± 0.11	9.31 ± 0.17	+1.71
SBM
S. mitis 1643	DAP (6 mg/kg)	6.96 ± 0.06	7.17 ± 0.34	−0.21
	DAP (8 mg/kg)	6.94 ± 0.13	7.46 ± 0.21	+0.52
	VAN	7.04 ± 0.05	6.5 ± 0.37	−0.54
	LZD	6.95 ± 0.18	6.61 ± 0.44	−0.34
	None (growth control)	6.58 ± 0.28	8.21 ± 0.13	+1.63
S. oralis 1647	DAP (6 mg/kg)	7.91 ± 0.30	7.42 ± 0.28	−0.49
	DAP (8 mg/kg)	7.17 ± 0.17	5.98 ± 0.23	−1.19
	VAN	7.31 ± 0.90	3.11 ± 1.1	−4.2
	LZD	6.90 ± 0.23	4.95 ± 0.30	−1.95
	None (growth control)	6.43 ± 0.24	7.95 ± 0.18	+1.52
S. oralis 1648	DAP (6 mg/kg)	7.12 ± 0.10	6.33 ± 0.18	−0.79
	DAP (8 mg/kg)	7.08 ± 0.23	6.62 ± 0.08	−0.46
	VAN	7.04 ± 0.05	3.26 ± 0.45	−3.78
	LZD	7.34 ± 0.08	5.81 ± 0.33	−1.53
	None (growth control)	7.32 ± 0.03	7.81 ± 0.16	+0.49
S. gordonii 1649	DAP (6 mg/kg)	7.41 ± 0.18	5.64 ± 0.27	−1.77
	DAP (8 mg/kg)	7.67 ± 0.06	6.39 ± 0.13	−1.28
	VAN	7.09 ± 0.13	3.92 ± 0.79	−3.17
	LZD	6.85 ± 0.21	6.14 ± 0.06	−0.71
	None (growth control)	7.23 ± 0.15	8.07 ± 0.21	+0.84

^aThe inoculum is given in log₁₀ CFU/g for SEVM or log₁₀ CFU/ml for SBM.

FIG 2.

Comparison of antimicrobial activities of daptomycin at 6 mg/kg (open circle) or 8 mg/kg (closed inverted triangle), vancomycin (open triangle), linezolid (closed square), and growth control (closed circle) in the SEVM against 4 VGS strains.

The SEVM model was selected as the initial and primary model to assess DAP activity against the VGS, as this was the most representative infection type to simulate with these organisms. However, when DAP activity data were not consistent with what had been demonstrated against staphylococci and enterococci in the same model, the SBM was then utilized to eliminate potential inference from the protein components of the SEVM.

The SBM model utilized the same regimens except for using the free (unbound) concentration of the antibiotics. Bacterial reduction over the experimental period is summarized in Table 4. DAP regimens achieved bacterial kill levels similar to those seen in the SEVM, with all strains having 1 to 2 log₁₀ CFU/ml reduction at 72 h, except for 8 mg/kg DAP against strain 1643 (0.52 log₁₀ CFU/ml increase). VAN achieved a ≥3 log₁₀ CFU/ml reduction for all strains except for strain 1643 (0.54 log₁₀ CFU/ml reduction). LZD achieved less than a 2 log₁₀ CFU/ml reduction for all strains.

Overall, VAN regimens (P ≤ 0.05) were consistently better than LZD regimens (P ≤ 0.05) compared to growth control against all strains in both the SEVM and SBM. DAP displayed significant activity only against strain 1649 in both models and strain 1648 in the SBM (P ≤ 0.05).

Fractional inhibitory concentration.

All RIF-DAP combinations showed indifference or additive activity with a FIC ranging from 0.75 to 2 for all strains. For GEN-DAP combinations, indifference or additive activity was observed for strains 1643 and 1647 with a FIC ranging from 0.625 to 1.06. Strain 1648 demonstrated antagonism with a FIC of 8.0625, and strain 1649 displayed synergy with a FIC of 0.3125.

Kill curves.

Due to the continued regrowth noted in the DAP studies despite removal of protein, additional time-kill studies were conducted to determine whether increased drug concentration could increase killing activity.

The KC results are displayed in Fig. 3. Of note, in some KCs, as the drug concentration increases, more bacteria were initially killed, followed by significant regrowth. In contrast, lower drug concentrations achieved some kill, but the curve was more static over the experimental period. Even drug concentrations at 8 times the MIC showed significant regrowth by 24 h. Regardless of protein content (albumin or PHS) or broth alone, the levels of bacteria killed were similar for all testing conditions.

FIG 3.

Activity of various concentrations of DAP in time-kill curve experiments evaluating the effect of protein (albumin and pooled human serum) on killing of bacteria. The testing condition and organism tested are indicated by a letter and number. Testing conditions were as follows: free (unbound) concentrations of DAP tested alone in broth (A), total concentrations of DAP tested in broth and albumin (B), and total concentrations of DAP tested in broth and pooled human serum (PHS) (C). Organisms tested were as follows: S. mitis 1643 (1), S. oralis 1647 (2), S. oralis 1648 (3), and S. gordonii 1649 (4). Regimens included growth control (closed circle), therapeutic DAP (6 mg/kg/day) (open circle), 0.5× MIC (closed inverted triangle), 1× MIC (open triangle), 2× MIC (closed square), 4× MIC (open square), and 8× MIC (closed diamond). Free MICs to determine DAP testing concentrations were 1 μg/ml for strains 1643 and 1648 and 2 μg/ml for strains 1647 and 1649, and total MICs were 256 μg/ml for strains 1643, 1647, and 1649 and 512 μg/ml for strain 1648.

Resistance.

All DAP postexposure strains from the SEVM and SBM displayed growth on DAP-containing plates at 4×, 8×, and 16× the preexposure strain MIC for all organisms. Subsequent susceptibility determinations yielded postexposure MICs of >256 μg/ml for all strains.

Resistance characterization.

Determination of genetic relatedness in the pre- and postexposure strains by RFLP and PFGE yielded identical patterns with strains 1643 and 5064, closely related patterns with strains 1647 and 5065 and 1648 and 5066, and uniquely different patterns between strains 1649 and 5067 (Fig. 4 and 5).

FIG 4.

Genetic relatedness of pre- and postexposure isolates by PFGE by single digestion for isolates 1643 and 5064 (S. mitis), isolates 1648 and 5066 and isolates 1647 and 5065 (S. oralis) and isolates 1649 and 5067 (S. gordonii).

FIG 5.

Genetic relatedness of pre- and postexposure isolates 1648 and 5066 (S. oralis) by PFGE by single and double digestion.

The heterogeneity screen for DAP resistance by modified DAP macro-Etest or plating on drug-containing media was positive for all of the preexposure strains by 48 h. The additional modified macro-Etest assays performed were negative for S. oralis 1647 and 1648 for both VAN and LZD at 48 h, while strain 1647 was positive for RIF heterogeneity. Colonies that appeared within the ellipse of the modified DAP macro-Etest were tested for their DAP MIC value by Etest, which resulted in a value 170 times the baseline MIC. Heterogeneity screening by plating on DAP-containing media yielded results similar to those of the modified DAP macro-Etest (presence of colonies on drug-containing plates at 4 to 64 μg/ml). Colonies that grew at 64 μg/ml from S. oralis 1647 and 1648 were tested for DAP MIC by Etest and demonstrated the same increase in MIC as seen with the modified DAP macro-Etest.

The resistant phenotype was maintained under selective pressure when serially passaged on DAP-containing plates; however, resistance was lost in 7 to 13 serial passages in the absence of DAP.

Cross-resistance, defined as a greater than three doubling dilution increase in the MIC of the postexposure strain compared to the preexposure strain, was not observed with any compound tested other than DAP (i.e., oxacillin, VAN, etc.).

The supernatant of cultures containing the highly DAP-resistant S. oralis 5065 and various concentrations of DAP was able to effectively inhibit the growth of S. pneumoniae in the DAP microbioassay at all test concentrations and time points without significant changes in the zone of inhibition (reduction in zone diameter of 4.9 to 5.9 mm compared to 3.4 to 4.1 mm in the no-organism control). HPLC analysis demonstrated that 47.9 to 62.1% of the original quantity of DAP remained after a 24-h period.

DISCUSSION

VGS are a clinically important group of Gram-positive pathogens often associated with infective endocarditis, bloodstream infections, and systemic diseases, particularly in immunocompromised patients. These commensal organisms are part of the upper respiratory and intestinal flora but predominantly reside in the oral cavity. With alternation or damage to the mucosal barrier, transient bacteremias may develop, potentially leading to clinically significant bloodstream infections which have been reported to occur in up to 24% of patients (38). Therefore, VGS bacteremia may lead to more serious complications, as these organisms have been associated with the highest percentage of native valve endocarditis (30 to 40%) compared to S. aureus (10 to 27%) (39).

Historically, penicillin has been the treatment of choice due to the inherent susceptibility of these organisms to most β-lactams; however, resistance to penicillins, as well as other β-lactams, macrolides, and fluoroquinolones, is increasing (4, 6, 7, 9, 12). Penicillin and macrolide resistance rates as high as ≥50% have been reported for VGS in some studies (4, 7, 40). Low levels of resistance to fluoroquinolones have been noted in most countries, with rates less than 3% (9, 41). However, this resistance rate increases significantly in cancer patients with previous exposure to fluoroquinolones, with resistance increasing to up to 64% for ciprofloxacin against S. mitis (11). Resistance to various antibiotic classes is often strain dependent, with S. mitis, S. oralis, and S. sanguinis tending to be the least susceptible.

DAP is a potent antibiotic with activity against multidrug-resistant Gram-positive organisms. Although species within the VGS are not listed in the Cubicin (DAP) package insert, the CLSI has recommended susceptibility interpretive criteria for Streptococcus viridans group at a broth microdilution MIC value of ≤1 μg/ml (28). Using CLSI breakpoints as a guideline, two of the four VGS strains (1647 and 1649) demonstrated baseline DAP MIC values in excess of the susceptibility breakpoint. The DAP surveillance program has monitored DAP susceptibilities to various Gram-positive isolates since 2002, which has consistently demonstrated >99% susceptibility of all VGS strains (20, 41). The MIC₉₀s for these organisms are most commonly reported to be 0.5 to 1 μg/ml with a MIC range of 0.06 to 2 μg/ml for the wild type. In vitro studies evaluating the activity of DAP against various streptococci, including VGS, have shown significant kill without development of resistance. There is only a single report for an in vitro peritoneal dialysate model where DAP at 10 mg/liter against Streptococcus sanguis (S. sanguinis) displayed more static activity with only a 2-log₁₀ CFU reduction compared to >3 log₁₀ CFU reduction with cefazolin and VAN (42). In addition to reduced activity of DAP, regrowth occurred in the model by 24 h with the authors attributing this to the possibility of inadequate calcium supplementation within the growth media despite noting that the calcium concentration was greater than 50 mg/liter.

In our study, DAP regimens failed to achieve bactericidal activity (over a 24-h period) in the in vitro SEVM, even at high doses (8 mg/kg/day). VAN treatment more consistently reduced the bacterial burden and to a larger extent than LZD or DAP did, resulting in a 1.6- to >3-log₁₀-unit reduction compared to a 1- to 2-log₁₀-unit reduction for LZD (Table 4 and Fig. 2). High-level daptomycin resistance (HLDR) was isolated postexposure, from all strains, in all DAP treatment groups (except 6 mg/kg daptomycin against strain 1649). DAP MIC values of the postexposure strains were greater than 16× the preexposure MIC (>256 μg/ml), demonstrating a rapid development of high-level resistance/selection by 24 h despite drug concentrations as high as 8 times the MIC. Similar activity was noted in the SBM utilizing free (unbound) drug concentrations, thereby eliminating the potential impact of protein as a confounding factor for ineffectiveness. Previous work has also demonstrated similar activity between SEVM, SBM, and KCs with DAP utilizing total drug concentrations in models containing protein versus free concentrations in models not containing protein (30, 33).

The results of this study, using the same models, are unique compared to similar treatment regimens against methicillin-resistant S. aureus, vancomycin-resistant Enterococcus faecium, and glycopeptide-intermediate S. aureus. In those studies, bactericidal activity was achieved and maintained over a 24-h period, and no resistance was detected postexposure (30, 33).

The presence of DAP tolerance in S. aureus has been widely studied and not detected in more than 900 isolates tested. All S. aureus isolates tested so far demonstrate an MBC/MIC ratio of 1 or 2, indicating rapid bactericidal activity of DAP (43–45). The DAP MBC/MIC ratios varied among the four VGS strains in this study (from 4 to 16) and were higher than the values seen with S. aureus. If we apply a commonly used definition of antibiotic tolerance (MBC/MIC ratio of ≥32 or MBC/MIC ratio of ≥16 with an MBC in the nonsusceptible/resistant range), strain 1643 would be considered DAP tolerant (46). DAP tolerance in VGS has not been previously described. Despite three of the four organisms not demonstrating tolerance, the elevated MBC/MIC ratio seems to be unique to these organisms. Tolerance, in addition to an elevated MBC/MIC ratio, has previously been demonstrated in several other drug classes, including penicillins, macrolides, lincosamides, fluoroquinolones, and glycopeptides (47–50), although it was not seen in the organisms used in this study. However, no mechanism of tolerance was determined in any study. Calcium concentrations of ≥50 mg/liter were included in all media/experiments due to DAP's calcium-dependent activity to ensure this was not a factor for the elevated MBC or MIC results. Additionally, the MBC/MIC ratios when drug was given in the presence of protein were also higher (from 2 to 8) than the ratio for S. aureus, although by definition, it was not DAP tolerant. All protein-containing experiments utilized total DAP concentrations, appropriately supplemented with calcium. Previous protein-containing experiments utilizing total DAP concentrations have been comparable to nonprotein experiments utilizing free (unbound) DAP concentrations (30, 33). Regardless of protein content, the MBC/MIC ratios were similar across all tested media; consequently, the inclusion of protein was not believed to impact the elevated ratios obtained.

The results of the KC experiments correspond well with those of the pharmacokinetic/pharmacodynamic models in that there was some level of initial killing and then regrowth to baseline CFU levels, an overall static effect at up to 8× the MIC value (Fig. 3). Isolates recovered at the 24-h time point had significantly elevated DAP MIC values. These results make sense in light of the fact that all of the preexposure strains were found to contain a DAP-resistant subpopulation as previously mentioned. This population of cells (∼0.1 to 1%) is most likely able to thrive under selective pressure and grow to baseline levels, while the majority of the susceptible cells are killed within 8 h.

Garcia-de-la-Maria and colleagues provide supporting evidence that this in vitro phenomenon of HLDR occurs rapidly in a study utilizing multiple pharmacodynamic models against numerous strains (23). Additionally, this study demonstrates that despite utilizing significant DAP concentrations, either free or total, of up to 8× MIC, only static activity was achieved. The addition of gentamicin resulted in enhanced activity and prevented the development of high-level resistance in several of the isolates studied. However, the clinical implications for the development of resistance seen in vitro are not currently known. Since the approval of DAP over 10 years ago, there have been limited published cases of clinical failure and/or resistance in enterococci and staphylococci but no reports regarding streptococci until recently (51, 52). The first case described was a clinical failure attributed to breakthrough bacteremia with DAP-resistant Streptococcus anginosus from a patient with recurrent methicillin-resistant S. aureus osteomyelitis that had been previously unsuccessfully treated with VAN; the patient was then switched to DAP treatment at 6 mg/kg/day. After 15 days of DAP therapy, this patient presented in septic shock with positive blood cultures for S. anginosus with a DAP MIC of 4 μg/ml (51).

The recent reporting of this case is important. However, a baseline isolate was not saved; therefore, it cannot be confirmed that this was a case of resistance development or an infection attributed to an organism with a DAP MIC slightly above those seen in the wild-type population. The moderately elevated MIC pales in comparison to the DAP MICs seen in this study and the study of Garcia-de-la-Maria et al. (23), as well as the rapidity of resistance development that was demonstrated. Additional clinical data pertaining to the treatment of these types of organisms with DAP will be essential in determining whether or not there is a clinical risk if DAP is used off label to treat VGS. Another recent study evaluated outcomes of VGS treated with daptomycin (52). Out of 9,105 patients, only 99 (1.1%) had VGS, for which DAP was utilized as the first-line agent in 21 patients (21%). Only 33 patients received DAP monotherapy, and the assessment of clinical outcomes was 73% success, 3% failure, and 24% nonevaluable. One patient, determined to be a clinical failure, was classified as such based on culture of a DAP-resistant isolate (DAP MIC of 2 μg/ml); a reason for failure was not specified for the other patients.

If these organisms can develop clinically and other therapeutic options are not available, combination therapy may be a potential solution to restrict the growth of the DAP-resistant subpopulation. However, the combinations examined in this study had disappointing results. The effects of the different antimicrobial combinations were strain/species specific, and the combinations may not be effective when applied to a larger collection of VGS. The only synergistic combination noted was GEN plus DAP against strain 1649, which should be investigated further. Additional combination therapy, in particular DAP with β-lactam agents, should be considered and investigated, as current literature indicates that daptomycin in combination with a β-lactam against staphylococci and enterococci exhibits enhanced activity and decreased resistance selection (53, 54).

Conclusion.

Certain VGS have the ability to rapidly develop HLDR after drug exposure, but the clinical relevance is lacking. Indeed, after more than 10 years of clinical use, surveillance continues to show very low resistance for VGS; however, a recent single clinical failure with a daptomycin-resistant strain of S. anginosus has been documented. Therefore, further studies are warranted to explore treatment options to prevent or minimize resistance development, particularly daptomycin in combination with β-lactams, given the evidence seen in staphylococci and enterococci. Additionally, studies should be conducted to elicit the mechanism of DAP resistance and to determine the clinical significance of this phenomenon.