https://orcid.org/0000-0002-7588-1733
Abstract
To explore whether linezolid/daptomycin combinations can restrict Staphylococcus aureus resistance and if this restriction is associated with changes in the mutant prevention concentrations (MPCs) of the antibiotics in combination, the enrichment of resistant mutants was studied in an in vitro dynamic model.
Two MRSA strains, vancomycin-intermediate resistant ATCC 700699 and vancomycin-susceptible 2061 (both susceptible to linezolid and daptomycin), and their linezolid-resistant mutants selected by passaging on antibiotic-containing medium were used in the study. MPCs of antibiotics in combination were determined at a linezolid-to-daptomycin concentration ratio (1:2) that corresponds to the ratio of 24 h AUCs (AUC24s) actually used in the pharmacokinetic simulations. Each S. aureus strain was supplemented with respective linezolid-resistant mutants (mutation frequency 10−8) and treated with twice-daily linezolid and once-daily daptomycin, alone and in combination, simulated at therapeutic and sub-therapeutic AUC24s.
Numbers of linezolid-resistant mutants increased at therapeutic and sub-therapeutic AUC24s, whereas daptomycin-resistant mutants were enriched only at sub-therapeutic AUC24 in single drug treatments. Linezolid/daptomycin combinations prevented the enrichment of linezolid-resistant S. aureus and restricted the enrichment of daptomycin-resistant mutants. The pronounced anti-mutant effects of the combinations were attributed to lengthening the time above MPC of both linezolid and daptomycin as their MPCs were lowered.
The present study suggests that (i) the inhibition of S. aureus resistant mutants using linezolid/daptomycin combinations can be predicted by MPCs determined at pharmacokinetically derived antibiotic concentration ratios and (ii) T>MPC is a reliable predictor of the anti-mutant efficacy of antibiotic combinations as studied using in vitro dynamic models.
Introduction
Global increases in antimicrobial resistance among common bacterial pathogens, including Staphylococcus aureus, are threatening the successful use of antibiotics such as linezolid. Despite the low probability of pre-existing bacterial resistance to linezolid,1,2 linezolid-resistant S. aureus indeed have been detected during treatment, with the first case reports appearing shortly after the introduction of the antibiotic into clinical practice.3–11 Our studies with linezolid-exposed S. aureus in an in vitro dynamic model have shown that ‘anti-mutant’ ratios of the 24 h AUC (AUC24) to the MIC (AUC24/MIC) might or might not be attainable at clinical doses.12,13 In the latter case, combination of the oxazolidinone with other antistaphylococcal agents could be useful. Daptomycin is a possible candidate, as data from numerous clinical case reports14–22 document daptomycin-resistant mutants emerging during treatment despite its high antistaphylococcal activity. For this reason, lipopeptide therapeutic AUC24/MIC ratios may not be ‘anti-mutant’, especially at a low dose (4 mg/kg). Therefore, combinations of both antibiotics are of interest as a means of limiting or overcoming antimicrobial resistance.
To our knowledge, the present work is one of a few studies focusing on the impact of linezolid/daptomycin combinations on the anti-mutant effect. Although the pharmacodynamics of this combination versus S. aureus has been studied previously in vitro23–28 and in vivo,29 when evaluated, the development of resistance to linezolid or daptomycin was not observed during simulated single or combined treatments.24–26 Thus, no conclusions can be suggested about linezolid/daptomycin interactions in relation to any anti-mutant effect.
The anti-mutant efficacy of linezolid- or daptomycin-containing combinations has been investigated in our previous studies with S. aureus exposed to linezolid in combination with rifampicin or gentamicin and daptomycin in combination with rifampicin.30–32 The observed enhancement of anti-mutant effects by antibiotic combinations was attributed to lengthening times when antibiotic concentration exceeded the mutant prevention concentration (MPC) as a result of lowering the MPCs of each antibiotic in the presence of the other within the dosing interval (T>MPC).30–32
To explore if combinations of linezolid with daptomycin can restrict S. aureus resistance and whether the possible restriction is associated with changes in MPCs, the enrichment of linezolid- and daptomycin-resistant mutants was studied by simulating single and combined antibiotic treatments at both therapeutic AUC24s (240 mg×h/L for a twice-daily 600 mg dose of linezolid33 and 480 mg×h/L for a once-daily 4 mg/kg dose of daptomycin34) and sub-therapeutic AUC24s at half the therapeutic values. As referenced above,30–32 MPCs of one antibiotic in the presence of the other were determined at a pharmacokinetically derived linezolid-to-daptomycin concentration ratio, which was equal to the ratio of the AUC24 of linezolid to the AUC24 of daptomycin used in pharmacokinetic simulations of the combined treatments.
Materials and methods
Antimicrobial agents, bacterial strains and susceptibility testing
Linezolid and daptomycin powders were purchased from Acros Organics (USA, Fair Lawn, NJ, USA).
Two MRSA strains were used in the study: well-characterized vancomycin-intermediate resistant ATCC 700699 (Mu50)35 and vancomycin-susceptible clinical isolate 2061. Both strains were susceptible to linezolid and daptomycin, with an MIC of the oxazolidinone of 2 mg/L and an MIC of the lipopeptide of 0.5 and 0.25 mg/L for S. aureus ATCC 700699 and 2061, respectively. Linezolid-resistant mutants (RMs; MIC of linezolid of 8 mg/L) of each strain were selected by passaging on antibiotic-containing medium according to a previously described procedure.12 Each S. aureus strain was supplemented with respective RMs to achieve a mutation frequency of 10−8 [RM content of one cell per 108 cfu of susceptible cells in 1 mL of Mueller–Hinton broth (MHB)]. This mixture was used in all MPC determinations and pharmacodynamic experiments.
MPC determinations
The MPCs of antibiotics alone and in the presence of each other were determined as described elsewhere.30 As in previous studies with linezolid/rifampicin and linezolid/gentamicin combinations,30–32 linezolid MPC in the presence of daptomycin, and daptomycin MPC in the presence of linezolid were determined at concentration ratios equal to the respective ratio of the antibiotic’s therapeutic AUC24s used in subsequent pharmacokinetic simulations (240 mg×h/L for the twice-daily 600 mg dose of linezolid33 to 480 mg×h/L for the once-daily 4 mg/kg dose of daptomycin34—1:2).
Antibiotic dosing regimens and simulated pharmacokinetic profiles
Both single and combined treatments mimicked therapeutic dosing regimens: 600 mg twice-daily linezolid and 4 mg/kg once-daily daptomycin. The respective AUC24s (240 and 480 mg×h/L) used in pharmacokinetic simulations corresponded to those reported in human studies.33,34 Dosing regimens of linezolid and daptomycin given alone were designated as L240 and D480, respectively, and in combination as L240+D480. Along with the therapeutic AUC24s, sub-therapeutic AUC24s at half of the therapeutic values were simulated with linezolid alone (AUC24 120 mg×h/L, regimen L120) and daptomycin alone (AUC24 240 mg×h/L, regimen D240) and in combination (regimen L120+D240). With all dosing regimens, a series of mono-exponential profiles that mimic twice-daily dosing of linezolid with a half-life of 6 h36 and once-daily dosing of daptomycin with a half-life of 9 h,34 alone or in combination, were simulated for 5 consecutive days.
In vitro dynamic model
A previously described dynamic model13 was used in simulations of single drug treatments with linezolid and daptomycin. Briefly, the model consisted of two connected flasks: one containing fresh MHB and the other with a magnetic stirrer, a central unit, with the same broth containing a bacterial culture plus antibiotic. Peristaltic pumps circulated fresh nutrient medium to and from the central 110 mL unit (initial 100 mL volume corrected by including additional 10 mL volume of the sampling system tubes) at a flow rate of 12.7 mL/h for linezolid or 8.5 mL/h for daptomycin.
To simulate combination treatments, i.e. to provide simultaneous mono-exponential elimination of linezolid and daptomycin with their inherent half-lives, the model was modified according to the Blaser and Zinner principle.37 The model was supplemented with an additional 54 mL flask with fresh MHB containing daptomycin at initial concentrations equal to those in the central unit. Peristaltic pumps circulated fresh nutrient medium to and antibiotic-containing medium (both linezolid and daptomycin) from the central unit at a flow rate of 12.7 mL/h, which corresponds to the antibiotic with the shorter half-life, i.e. linezolid. To compensate for a too-rapid decrease in concentrations of the antibiotic with the longer half-life (daptomycin), peristaltic pumps circulated fresh medium to and daptomycin-containing medium from the additional flask to the central unit at a flow rate of 4.2 mL/h (12.7–8.5 mL/h).
The procedure used in the pharmacodynamic experiments is described elsewhere.13 The duration of each experiment was 120 h. Each experiment was performed at least in duplicate. Antibiotic dosing and sampling the central unit of the dynamic model were processed automatically using computer-assisted systems.
Quantitation of the antimicrobial effects on the resistant subpopulations of S. aureus
To determine viable counts of linezolid- and daptomycin-resistant S. aureus mutants, the central unit of the model was multiply sampled throughout the observation period (120 h) and the samples were serially diluted, if necessary, and plated on Mueller–Hinton agar (‘MHA’) with 2×MIC and 4×MIC of linezolid. The inoculated plates were incubated for up to 72 h at 37°C and screened visually for growth. The lower limit of detection was 10 cfu/mL (equivalent to at least one colony per plate).
Time courses of mutants resistant to 2×MIC of linezolid or daptomycin were characterized by the area under the bacterial mutant concentration–time curve [AUBCM(L) or AUBCM(D), respectively]38 calculated from time 0 to 120 h after the start of treatment and corrected for the area under the lower limit of detection over the same time interval. The correlation of AUBCM with resistant mutant killing is inversely related; the greater the anti-mutant effect, the lower the AUBCMs.
Results
MPCs of linezolid and daptomycin alone and in combination
MPCs were assessed for linezolid and daptomycin alone and in combination (concentration ratio 1:2). Numbers of surviving S. aureus cells decreased systematically with increasing antibiotic concentrations in agar plates; the combination plots were shifted to the left along the abscissa, with the shift being more pronounced with linezolid than daptomycin (Figure 1). As a result, the estimated MPCs of linezolid or daptomycin in combination were lower than the MPCs observed with the single agents. Under the influence of daptomycin the MPCs of linezolid decreased from 10 to 3 mg/L (3.3-fold, S. aureus ATCC 700699) and to 4 mg/L (2.5-fold, S. aureus 2061), whereas under the influence of linezolid the MPCs of daptomycin decreased from 14 to 6 mg/L (2.3-fold) and from 10 to 8 mg/L (1.3-fold).
Determination of MPC of linezolid and daptomycin alone and in combination.
Antibiotic pharmacodynamics with resistant S. aureus mutants
Simulated pharmacokinetic profiles of linezolid alone and in combination with daptomycin and the respective time courses of S. aureus ATCC 700699 mutants resistant to 2×MIC of linezolid are shown in Figure 2. As seen in the figure, with S. aureus ATCC 700699 exposed to linezolid alone at therapeutic (L240) and sub-therapeutic (L120) exposures, numbers of linezolid-resistant mutants increased starting from 24 and 0 h and reaching a level of ca. 105 and 108 cfu/mL, respectively, by the end of the observation period. Unlike single drug treatments, the combined use of linezolid with daptomycin (regimens L240+D480 and L120+D240) completely suppressed linezolid-resistant mutants throughout the observation period. Similarly, the enhanced anti-mutant effect of combined therapy against linezolid-resistant mutants was observed both at therapeutic and sub-therapeutic exposures in experiments with S. aureus 2061 (Figure S1, available as Supplementary data at JAC Online). Similar, but less pronounced, bacterial growth on agar plates with 4×MIC of linezolid was observed for both S. aureus strains (data not shown).
Simulated pharmacokinetics of linezolid and time courses of S. aureus ATCC 700699 mutants resistant to 2×MIC of the antibiotic.
The emergence of daptomycin-resistant S. aureus mutants of both strains exposed to daptomycin monotherapy was less pronounced than with linezolid, especially with the therapeutic regimen. With S. aureus ATCC 700699, the selection of daptomycin-resistant mutants was observed starting from 48 h of the experiment with both regimens, with a slight (D480) or more rapid increase (D240) in mutant numbers towards the end of the simulated treatment (Figure 3). However, combined treatments resulted in complete suppression (regimen L240+D480) or restriction (regimen L120+D240) of daptomycin-resistant organisms. With S. aureus 2061, daptomycin monotherapy at therapeutic exposure (D480) protected against mutant enrichment; therefore, addition of linezolid did not significantly influence the anti-mutant effect of daptomycin. At sub-therapeutic exposure (D240), slight growth of daptomycin-resistant mutants was observed, but it was suppressed when linezolid was added (L120+D240) (Figure S2). With both S. aureus strains, bacterial growth was not observed on agar plates containing 4×MIC of daptomycin (data not shown).
Simulated pharmacokinetics of daptomycin and time courses of S. aureus ATCC 700699 mutants resistant to 2×MIC of the antibiotic.
Enhancement of the anti-mutant effect of linezolid expressed as AUBCM(L) (the greater the anti-mutant effect, the lower the AUBCMs) in the presence of daptomycin was consistent with increased T>MPC. (Figures 4a and b and 5a and b). For example, in experiments with S. aureus ATCC 700699, AUBCM(L)s declined dramatically from 256 to 14 (log cfu/mL)×h (regimen L240+D480) and from 526 to 0 (log cfu/mL)×h (regimen L120+D240) along with the lengthening of T>MPC from 47% to 100% and from 0% to 84%, respectively. Change in AUBCM(D)s under the influence of linezolid was less pronounced than the corresponding change in AUBCM(L)s (Figures 4c and d and 5c and d). Again, the enhancement of daptomycin’s anti-mutant effect in the presence of linezolid with both S. aureus strains was consistent with T>MPC. For example, with S. aureus ATCC 700699, when AUBCM(D) decreased from 97 to 24 (log cfu/mL)×h at therapeutic exposures and from 247 to 66 (log cfu/mL)×h at sub-therapeutic exposures, it could be attributed to respective increases in T>MPC from 65% to 100% and from 27% to 73% of the dosing intervals. With the clinical S. aureus isolate, AUBCM(D) for sub-therapeutic regimens decreased from 149 to 47 (log cfu/mL)×h and T>MPC increased from 44% to 57%. Of note, with this strain at the therapeutic exposure (D480) the high value of T>MPC (82%) is responsible for the anti-mutant effect of daptomycin alone; the addition of linezolid could not further lower the AUBCM(D) and T>MPC was 94% of the dosing interval.
Enrichment of S. aureus ATCC 700699 mutants resistant to 2×MIC of linezolid or daptomycin expressed as AUBCM and the respective values of T>MPC.
Discussion
In the present study with two strains of S. aureus, MPCs of linezolid and daptomycin in combination were determined at a linezolid-to-daptomycin concentration ratio of 1:2, which corresponds to the therapeutic antibiotic AUC24 ratio used in the pharmacokinetic simulations. With both strains, the MPC of each antibiotic in combination decreased and this led to increases in the respective T>MPCs of daptomycin and linezolid in the combined treatments. Simultaneously with increased T>MPCs for each antibacterial in the presence of the second agent, the enrichment of resistant mutants was restricted or fully suppressed as reflected by decreased AUBCMs. Obviously, this demonstrates the protective effects against the development of S. aureus resistance of lengthening the times above MPC.
The predictive anti-mutant potential of T>MPC against both Gram-positive and Gram-negative organisms has been confirmed previously in in vitro and in vivo pharmacodynamic experiments with several antibacterial classes (fluoroquinolones,39–43 β-lactams,44,45 fosfomycin46 and linezolid47,48). Also, the reliability of T>MPC as a predictor of anti-mutant effects of antibiotic combinations against staphylococci was confirmed in a study with linezolid plus rifampicin, where the approach for determining MPCs of combinations at pharmacokinetically derived antibiotic concentration ratios was first introduced.13 A subsequent study with linezolid plus gentamicin supported these findings.31 In these combination studies, T>MPCs of each antibiotic lengthened and linezolid-resistant mutants were not selected, whereas they were observed when the oxazolidinone was used alone. Similarly, amplification of gentamicin- and rifampicin-resistant mutants observed in mono-treatments was restricted or suppressed with the addition of linezolid.
The enhancement of the anti-mutant efficacy of linezolid and daptomycin in the presence of each other was not observed in another staphylococcal study with this antibiotic combination.49 The expansion of daptomycin resistance in the presence of linezolid was observed. This might have resulted from specific resistance mechanisms that were selected in the presence of these two antibiotics under static conditions; daptomycin concentrations increased stepwise, while linezolid concentrations were maintained at sub-MIC levels. In contrast to our pharmacodynamic study where both antibiotic concentrations were always changing, this might influence the mechanisms by which resistance develops.
In the current study with S. aureus, T>MPC was confirmed as a predictor of anti-mutant efficacy of linezolid and daptomycin combinations. A clear concordance between AUBCM(L) or AUBCM(D) and T>MPC was observed; increases in T>MPCs of linezolid and daptomycin were associated with decreased AUBCMs (Figures 4 and 5). Thus, the anti-mutant effects of linezolid/daptomycin combinations can be predicted using T>MPC. However, future experiments should be designed to establish the T>MPC–resistance relationship over a wide range of T>MPCs with linezolid/daptomycin and other antibiotic combinations. This approach would likely provide MPC breakpoints for anti-mutant efficacy with several antibiotics in combination and could support the appropriate clinical use of these agents in the era of increased antimicrobial resistance.
Enrichment of S. aureus 2061 mutants resistant to 2×MIC of linezolid or daptomycin expressed as AUBCM and the respective values of T>MPC.
It is worth noting that our study has some limitations. First, we did not perform a genetic analysis of mutant cells that proliferated during experiments to reveal the mechanisms underlying resistance to linezolid and to daptomycin. Understanding the genetic determinants of resistance could be helpful in exploring the specific nature of the antimicrobial resistance observed in our experiments. Second, the use of only two S. aureus strains limits the potential clinical relevance of our findings.
Conclusions
The present study suggests that (i) the inhibition of S. aureus resistant mutants using linezolid/daptomycin combinations can be predicted by MPCs determined at pharmacokinetically derived antibiotic concentration ratios and (ii) T>MPC is a reliable predictor of the anti-mutant efficacy of antibiotic combinations as studied using in vitro dynamic models.
Acknowledgements
This study was presented in part at the Twenty-Ninth European Congress of Clinical Microbiology and Infectious Diseases, Amsterdam, The Netherlands, 2019 (Abstract no. P2116) and at the Thirtieth European Congress of Clinical Microbiology and Infectious Diseases, 2020 (Abstract no. 3357).
Funding
This study performed at the Department of Pharmacokinetics & Pharmacodynamics, Gause Institute of New Antibiotics, was supported by a grant from the Russian Science Foundation (no. 18-15-00433).
Transparency declarations
None to declare.
Supplementary data
Figures S1 and S2 are available as Supplementary data at JAC Online.
References
Author notes
This author has passed away.
