|Home | About | Journals | Submit | Contact Us | Français|
Since the currently approved dose of daptomycin (6 mg/kg of body weight/day) has been associated with clinical failures and resistance development, higher doses for some difficult-to-treat infections are being proposed. We studied the efficacy of daptomycin at high doses (equivalent to 10 mg/kg/day in humans) and compared it to that of reference and alternative treatments in a model of foreign-body infection with methicillin (meticillin)-resistant Staphylococcus aureus. In vitro studies were conducted with bacteria in the log and stationary phases. For the in vivo model, therapy with daptomycin at 100 mg/kg/day, vancomycin at 50 mg/kg/12 h, rifampin (rifampicin) at 25 mg/kg/12 h, or linezolid at 35 mg/kg/12 h was administered for 7 days. Antibiotic efficacy was evaluated using either bacteria from tissue cage fluids or those attached to coverslips. We screened for the emergence of linezolid- and rifampin-resistant strains and analyzed the surviving population from the daptomycin-treated group. Only daptomycin was bactericidal in both the log- and stationary-phase studies. Daptomycin (decrease in the log number of CFU per milliliter of tissue cage fluid, 2.57) and rifampin (decrease, 2.6 log CFU/ml) were better (P < 0.05) than vancomycin (decrease, 1.1 log CFU/ml) and linezolid (decrease, 0.9 log CFU/ml) in the animal model. Rifampin-resistant strains appeared in 60% of cases, whereas no linezolid resistance emerged. No daptomycin-resistant subpopulations were detected at frequencies of 10−7 or higher. In conclusion, daptomycin at high doses proved to be as effective as rifampin, and the two were the most active therapies for this experimental foreign-body infection. These high doses ensured a profile of safety from the development of resistance.
Daptomycin is a lipopeptide drug with bactericidal activity toward methicillin (meticillin)-resistant Staphylococcus aureus (MRSA) in a concentration-dependent manner (27, 32). It is currently approved for use at 4 mg/kg of body weight/day for skin and soft-tissue infections (1) and at 6 mg/kg/day for bacteremia and right-side endocarditis (12, 14).
In recent years, reports of clinical failures and the emergence of resistant strains following daptomycin treatment have raised great concern (6, 18, 31). As a result, higher doses of daptomycin are being proposed as an alternative for some difficult-to-treat infections such as complicated bacteremia and endocarditis. Recently, doses of 10 mg/kg/day were studied using an in vitro model of staphylococcal endocarditis, with the results being promising in terms of efficacy and resistance prevention (25, 26). To date, clinical experience with the activity of the drug at doses higher than 6 mg/kg/day is limited (15, 28), whereas good safety and tolerance profiles for daptomycin at up to 12 mg/kg/day in volunteers have been reported (4).
Foreign-body infections are difficult to treat because of the presence of bacterial biofilm and tolerance to antibiotics (10, 39). MRSA is commonly involved in such infections, and daptomycin may be a promising drug (39). However, clinical experience in this area is again scarce, and the recommended doses are not clearly established, there being reports of clinical failures with doses of 4 to 6 mg/kg/day (13, 24).
The rat model of tissue cage infection is a well-standardized model that reasonably mimics human device infections (17, 20, 35). In this model, the efficacy of daptomycin has been partially studied (29, 34).
Taken together, the available experimental and clinical data seem to indicate that high doses of daptomycin are required in the setting of foreign-body infection.
In the present study, we aimed to test the efficacy of daptomycin at doses equivalent to 10 mg/kg/day in humans in a model of foreign-body infection with MRSA, comparing it with the efficacy of the current reference or main alternative treatments such as vancomycin, rifampin (rifampicin), and linezolid. We also sought to analyze the protection offered against the emergence of resistant strains.
For all the in vitro and in vivo studies, we used a MRSA strain (HUSA 304).
MICs and minimal bactericidal concentrations (MBCs) for bacteria in the log phase were determined by following standard recommendations (9).
MBCs for bacteria in the stationary phase were also determined. The methodology used has been reported previously and proved to be a reliable approach for correlating in vivo efficacy in the tissue cage model of infection (20, 38). MBCs were defined as the minimal concentration of antibiotic that was able to kill 99.9% of bacteria from the initial inoculum.
The methodology used for the log phase followed previously standardized recommendations (22), while the one used to obtain 24-h time-kill curves for the stationary phase has been described in detail elsewhere (20). The concentrations of antibiotics selected to establish time-kill curves for the log phase were those representing subinhibitory and clinically achievable levels greater than the MIC. Due to the tolerance of bacteria in the stationary phase to antibiotics, the drug concentrations used in the stationary-phase experiments were higher than those used in the log-phase experiments; the prefixed antibiotic concentrations were equivalent to peak and trough levels in tissue cage fluids (TCF).
Bactericidal activity was defined as that killing ≥3 log CFU/ml at 24 h with respect to the initial inoculum.
To evaluate daptomycin activity, the nutrient medium was supplemented with calcium (50 mg/liter; Sigma) in all the procedures.
The animal model had been previously approved by the Ethical Committee for Animal Experiments at the University of Barcelona.
Two Teflon tissue cages per animal with two coverslips each were subcutaneously implanted into male Wistar rats. After 3 weeks, TCF were checked for sterility and infected with 0.1 ml of a MRSA preparation (106 CFU/ml). At 72 h postinoculation (designated day 1), TCF were obtained to determine bacterial counts. Therapy was then started and administered intraperitoneally for 7 days; each therapeutic group included 10 animals. One and four days after the end of treatment (days 8 and 11, respectively), TCF were again obtained to determine bacterial counts. On day 11, animals were sacrificed and coverslips were removed to quantify adherent bacteria. The criterion for efficacy was defined as a decrease in bacterial counts in TCF between the beginning and end of treatment; it was thus evaluated twice, on days 8 and 11. The antibiotic efficacy against adherent bacteria from coverslips removed on day 11 was also evaluated.
We had previously standardized a model of chronic tissue cage infection (3 weeks old) with methicillin-susceptible S. aureus (20, 21), but for the present study we developed a model of earlier foreign-body infection (3 days old) with MRSA (38). Thus, prior to therapeutic experiments we analyzed the spontaneous evolution of MRSA infection. During the first week postinoculation, a decrease in bacterial counts with respect to the initial inoculum was observed; beyond the first week, the infection tended to be chronic and stable. The bacterial counts (mean log numbers of CFU per milliliter ± standard deviations [SD]) for TCF across the postinoculation period were as follows: 5.5 ± 0.5 (1st day), 5.3 ± 0.7 (2nd day), 5.5 ± 0.7 (3rd day), 5.5 ± 0.9 (4th day), 5.5 ± 0.3 (5th day), 5.5 ± 0.9 (6th day), 5.4 ± 1 (7th day), 6.4 ± 0.9 (8th day), 6.3 ± 0.8 (10th day), 6.2 ± 0.9 (14th day), and 6.5 ± 1 (21st day).
At 72 h after inoculation, we checked for the presence of macroscopic biofilm and quantified the mean bacterial population adhered to coverslips as 5 ± 0.9 log CFU/ml by following the methodology described in detail previously (20). In brief, the recovered coverslips were rinsed three times in phosphate-buffered saline, and they were then incubated in 1 ml of phosphate-buffered saline with trypsin and sonicated; the bacteria in the final fluid were quantified, and counts were recorded as log numbers of CFU per milliliter.
All the methodology used was described in detail previously (20, 21). We mainly adjusted the area under the concentration-time curve from 0 to 24 h (AUC0-24) for each antibiotic that showed similar results in animals (evaluated with either TCF or sera) and humans (evaluated with sera) (11, 33).
To test for equilibrium drug concentrations, samples of TCF were obtained on day 4 of treatment.
The main pharmacokinetic parameters for the dosages of linezolid, vancomycin, and rifampin have been reported previously by our group (20). The maximum concentrations (Cmax) of the drugs in sera and TCF were as follows: 38 and 17 μg/ml, respectively, for linezolid, 80 and 31 μg/ml for vancomycin, and 24 and 6.6 μg/ml for rifampin. The AUC0-24 values for the drugs in sera and TCF were as follows: 302 and 245 μg·h/ml for linezolid, 302 and 343 μg·h/ml for vancomycin, and 277 and 304 μg·h/ml for rifampin.
In the case of daptomycin, studies with several doses were performed to select the most appropriate one that was equivalent to the human dose of 10 mg/kg/day.
Daptomycin concentrations were determined by a bioassay method (7) with Micrococcus luteus ATCC 9341 (linearity of assay [r2], 0.99; lower detection limit, 2 μg/ml).
Animals were divided into therapeutic groups and received the following dosage: daptomycin, 100 mg/kg/day; linezolid, 35 mg/kg/12 h; vancomycin, 50 mg/kg/12 h; or rifampin, 25 mg/kg/12 h. Controls received no drug.
The screening of resistant strains from TCF and coverslips at the end of therapy was performed for linezolid and rifampin by using agar plates containing 4 and 1 μg/ml of the respective drugs.
TCF samples were checked on days 8 and 11, whereas coverslips were screened on day 11. In all cases, a sample of 100 μl from TCF or fluid from processed coverslips and 10-fold dilutions of 100-μl samples were inoculated onto the agar plates, and the plates were then incubated at 37°C for 48 h. Results were interpreted as positive (any macroscopic growth) or negative (no growth).
For the particular case of daptomycin, the above-mentioned screening was complemented by an analysis of the surviving populations from therapeutic experiments. The methodology followed previous recommendations (5); in brief, bacteria recovered from TCF or coverslips at the end of therapy were cultured overnight in Trypticase soy broth. They were then resuspended and adjusted to a concentration of 108 CFU/ml. Finally, a sample of 100 μl of the suspension and corresponding 10-fold dilutions were inoculated onto agar plates containing several daptomycin concentrations (1, 2, 3, 4, and 5 μg/ml). Any bacteria grown on these plates were consecutively passed onto drug-free agar plates for 5 days and then tested to determine the MIC. All the plates containing daptomycin were supplemented with calcium (50 mg/liter).
All bacterial counts are presented as log numbers of CFU per milliliter (means ± SD). Differences in bacterial counts for treated and untreated animals were evaluated for statistical significance by using analysis of variance. An unpaired Student t test with the Bonferroni correction was used to determine statistical significance. For all tests, differences were considered to be statistically significant when P values were <0.05.
The MICs and MBCs (in micrograms per milliliter) for bacteria in the logarithmic and stationary phases are shown in Table Table11.
Time-kill curves for the log phase showed bactericidal activity for all antibiotics except linezolid, and at concentrations achievable in clinical practice. Linezolid showed only a bacteriostatic effect.
Time-kill curves for the stationary phase showed bactericidal activity exclusively for daptomycin, although higher doses than those in the log-phase experiments were needed (24× MIC versus 4× MIC). Rifampin did not strictly achieve a bactericidal effect, although it was the second most active drug; we also noted that regrowth of resistant strains occurred in some experiments. Linezolid was the least effective drug against nongrowing bacteria.
Time-kill curves for clinically relevant concentrations achieving the highest levels of efficacy during both phases are shown in Fig. Fig.1.1. Note that concentrations used in the stationary-phase studies were higher than those used in the log-phase studies.
We performed pharmacokinetic studies with rats and several doses of daptomycin (45, 70, and 100 mg/kg/day) in order to select the most appropriate dosage that was equivalent to a human dose of 10 mg/kg/day.
The selected dose of daptomycin was 100 mg/kg/day. The main pharmacokinetic parameters for all dosages of daptomycin tested and their ratios to the MIC are shown in Table Table22.
A total of 50 animals (100 tissue cages) were used. Mean bacterial counts ± SD in TCF from the different therapeutic groups on day 1 were as follows: 5.44 ± 1.1 log CFU/ml for the daptomycin-treated group (n = 20 tissue cages), 5 ± 0.9 log CFU/ml for the linezolid-treated group (n = 20), 5.5 ± 0.9 log CFU/ml for the vancomycin-treated group (n = 20), 5.54 ± 1 log CFU/ml for the rifampin-treated group (n = 20), and 5.43 ± 0.8 log CFU/ml for controls (n = 15). There were no significant differences among the groups.
At the end of therapy, a double evaluation of efficacy was conducted by determining bacterial counts in TCF on days 8 and 11. The decreases in bacterial counts for each group are presented in Table Table33.
Results for all therapeutic groups were better than those for controls in both evaluations. Overall, daptomycin and rifampin were the most effective treatments, and they proved to be better than linezolid and vancomycin (P < 0.05) on both days. Daptomycin and rifampin achieved slightly greater efficacy on day 11 than on day 8, whereas vancomycin and linezolid exhibited less activity on day 11.
The efficacy against bacteria adhered to coverslips was evaluated on day 11. The bacterial counts (means ± SD) were as follows: 2.1 ± 1.2 log CFU/ml for the daptomycin-treated group, 3.1 ± 1.1 log CFU/ml for the linezolid-treated group, 2.9 ± 1.1 log CFU/ml for the vancomycin-treated group, 2.1 ± 1.3 log CFU/ml for the rifampin-treated group, and 5.2 ± 0.8 log CFU/ml for controls. Results for all therapeutic groups were better than those for controls (P < 0.05). The most effective treatments were again daptomycin and rifampin, and both were better than linezolid (P < 0.05) and vancomycin (P = 0.06).
At the end of therapy, resistant strains from samples of TCF or coverslips from the rifampin group were detected, but that was not the case for the daptomycin and linezolid groups. Rifampin-resistant strains were up in number between days 8 and 11 probably because selective pressure was maintained beyond the end of therapy and resistant strains continued growing. These results are presented in Table Table44.
Specific studies of survival populations from daptomycin experiments (no. of TCF samples = 8) showed that no resistant subpopulations in the presence of daptomycin at 1, 2, 3, 4, and 5 μg/ml were detected at frequencies of 10−7 or higher.
However, in some cases with high inocula of 108 CFU/ml, we did observe the growth of a bacterial haze on agar plates supplemented with the drug (1 to 5 μg/ml), and it was impossible to count the bacteria. Moreover, only one 10-fold dilution of inocula of 108 CFU/ml did not correspond to the reduction of bacterial counts, since no bacterial growth was detected when inocula of 107 CFU/ml were used. Studies with the wild-type strain HUSA 304 confirmed the same phenomenon.
These bacteria grown on plates containing daptomycin were recovered and passed consecutively onto drug-free agar plates over a period of 5 days; MICs were then determined again and were found to be the same as the initial value (1 μg/ml) in all cases. We thus confirmed that no changes in the MIC with respect to that for the initial strain occurred.
The appearance of clinical failures and resistant strains during daptomycin treatment has led to the suggestion that daptomycin be used at high doses for difficult-to-treat infections (6, 18). This study focused on the use of such doses against an experimental foreign-body infection with MRSA.
Daptomycin is rapidly bactericidal toward MRSA, and the findings of our in vitro studies were concordant with previous reports of good killing profiles against both growing and nongrowing bacteria, although higher concentrations for the stationary phase were needed (19). In contrast, the in vitro activities of vancomycin and rifampin were affected by the inoculum and the growth phase of bacteria; both were bactericidal exclusively against log-phase bacteria.
The contribution of daptomycin and its recommended dose in the setting of orthopedic prosthetic infections have not been definitively established (13, 24). The main role of combined therapies including rifampin is well known (36, 40); however, when MRSA strains are resistant to rifampin, antibiotic monotherapy may be the only available option. In such cases, frequent failures with vancomycin, usually considered to be the treatment of reference, or linezolid, the main alternative, can be anticipated (39). Although daptomycin that is active against nongrowing bacteria may be a promising drug in the treatment of foreign-body infection, its activity at high doses against MRSA has not been evaluated previously. Daptomycin remains active against nongrowing bacteria, although concentrations higher than those effective against growing bacteria are needed.
As regards the present study, there are a number of original findings about the activities of the antibiotics tested that require discussion. Daptomycin at high doses and rifampin were the best treatments, and particularly, we noted that daptomycin at a dose equivalent to 10 mg/kg/day proved to be better than vancomycin, an especially promising finding. On the other hand, the good in vivo efficacy of rifampin against device infection was again confirmed, despite the appearance of resistant strains (20, 36, 38). Based on the known good activity of rifampin as a comparator agent, the results obtained with high doses of daptomycin are also noteworthy.
Since daptomycin is bactericidal toward S. aureus in a concentration-dependent manner, the use of high doses seems to determine the final efficacy in our model (27), and in the same way, our results raise the matter that doses of up to 10 mg/kg/day may lead to even greater efficacy. The fact that we did not compare high doses of daptomycin with conventional ones may be considered a limitation of this study. However, we assumed that results from previous experimental studies and clinical data already support the use of high doses, and ethical reasons required us to restrict the number of animals (6, 25, 29). In this regard, daptomycin has been tested previously in a rat model of foreign-body infection at a dosage of 30 mg/kg/day against MRSA (34) and at 30 mg/kg/12 h against methicillin-susceptible S. aureus (29). The previous results would seem to suggest that the higher the dose of daptomycin, the better the outcome. In the first study, the dose of daptomycin was equivalent to 4 mg/kg/day in humans and daptomycin at this level was as effective as vancomycin; however, resistant subpopulations emerged. In the second study, daptomycin was significantly better than oxacillin and vancomycin but the equivalence to the dosage in humans was difficult to interpret because of the twice-daily administration. Since our aim was to update the stated high doses of daptomycin to that equivalent to 10 mg/kg/day in humans, we sought to define strictly the equivalence between rat and human dosages for once-daily administration in order to select the most appropriate dose. In this way, since the half-life of daptomycin in rats is shorter than that in humans, the AUC0-∞ was more accurate than other pharmacokinetic parameters (i.e., the Cmax and trough concentration) for making comparisons between rats and humans with once-daily administration.
A known limitation for the efficacy of daptomycin is the level of calcium, and this factor may be relevant in local environments. Although we did not determine calcium values for TCF, it has been reported previously that calcium levels in the tissue cage model are similar to those found in human serum (37).
Linezolid is currently used against orthopedic prosthetic infections, although clinical experience is limited (3). Linezolid achieved the least killing, as reflected in bacterial counts for TCF, and it was significantly less active than daptomycin, although not significantly worse than vancomycin. Similarly limited efficacy in a guinea pig model of tissue cage infection has been reported recently (2). We confirmed the protection against resistance reported previously for this antibiotic (16).
In recent years, great concern has arisen due to reports about the emergence of daptomycin-resistant strains of S. aureus during treatment (6, 18, 31). The use of daptomycin at high doses in an in vitro model of endocarditis caused by S. aureus proved to be effective in preventing any change in initial MICs (25, 26). Our in vivo results showed that no resistant strains appeared, even among bacteria embedded in biofilm from coverslips. The analysis of bacterial subpopulations showed the same frequencies of mutants for the wild-type strain and surviving bacteria from therapeutic experiments. However, we noted that when high inocula of 108 CFU/ml were studied, a haze of bacteria grew on agar plates containing daptomycin concentrations higher than the MIC. These bacteria have been previously termed “false mutants” (23, 30); in our study, we also confirmed the recovery of the initial MIC after bacteria were consecutively passed onto drug-free agar. This phenomenon may be an in vitro inoculum effect, although its potential relevance in vivo is still unknown.
Overall, daptomycin at high doses seems to be the most effective single option for device infection, especially against multidrug-resistant MRSA when limited therapeutic alternatives exist. In fact, the in vitro activity of antibiotics against nongrowing bacteria has been related previously to great efficacy against foreign-body infection (20, 36, 38); once again, our daptomycin results were in accordance with these findings. While daptomycin concentrations showing bactericidal activity are easily achieved in human serum (even at a conventional dosage of 6 mg/kg/day), lesser concentrations at local sites or the presence of tolerant strains in deep-seated infections may be responsible for suboptimal efficacy. In this regard, our results from testing high daptomycin concentrations locally available in TCF show good in vivo efficacy, as could be predicted given that concentrations were mostly above the in vitro MBC for the stationary phase.
In conclusion, daptomycin at high doses proved to be as effective as rifampin, and both were the most active therapies against MRSA for this foreign-body infection. These high doses of daptomycin ensured a profile of safety from the emergence of resistance.
This work was supported by a research grant from the Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III (FIS 08/014); grants from the Spanish Network for Research in Infectious Diseases (REIPI C03/14 and REIPI RD06/0008); and grants from Pfizer (Spain) and Novartis (Spain). O.M. was supported by a grant from the REIPI.
Published ahead of print on 27 July 2009.