Linezolid tissue penetration and serum activity against strains of methicillin-resistant Staphylococcus aureus with reduced vancomycin susceptibility in diabetic patients with foot infections

Objectives : Linezolid soft tissue penetration and serum antimicrobial activity were analysed in six patients with peripheral vascular disease and severe diabetic foot infections requiring surgical inter-vention. Methods : Blood draws (1, 3, 6, 9 and 12 h after initiation of a 1 h infusion) and a viable soft tissue sample at the site of infection were obtained in patients receiving linezolid (600 mg every 12 h) on the day of surgery. Concentrations of linezolid were determined by HPLC in both tissue (pre-treated with tissue lysis buffer) and serum. In addition, serum inhibitory and bactericidal activity (dilution titres 1:2–1:32) of linezolid was determined in these patients against strains of methicillin-resistant Staphylococcus aureus (MRSA) with reduced susceptibility to vancomycin (vancomycin MICs 5 2, 4, 8, 256 and > 256 mg/L). Results : Linezolid concentrations in tissue were found to be 51% (range, 18% to 78%) of simultaneous serum concentrations. Rapid (1 h) and prolonged (12 h) inhibitory activity (titres (cid:2) 1:2) was observed for linezolid against each of the study isolates. Furthermore, bactericidal activity (titres (cid:2) 1:2) was observed for at least 6 h (50% of the dosing interval) against four of these ﬁve strains. Conclusions : These ﬁndings suggest that linezolid could be effective in the treatment of multidrug-resistant MRSA even when concentrations at the infection site are diminished due to impaired blood ﬂow.


Introduction
The treatment of diabetic foot infections is complex and complicated by several risk factors that include peripheral vascular disease, autonomic and sensory neuropathy, and hyperglycaemia. 1 Antibiotic selection initially involves decisions concerning the severity of the infection, the route of therapy, the spectrum of organisms to be covered and potential adverse effects. 2 Another key to successful antibiotic therapy is achieving therapeutic drug concentrations at the site of infection. 3 When peripheral vascular disease is present, therapeutic antibiotic concentrations in infected tissues may not be achieved despite adequate serum levels. 4 An antibiotic regimen for diabetic foot infections should generally include an agent active against Staphylococcus aureus and b-haemolytic streptococci. More chronic wounds may need extended coverage that includes Gram-negative bacilli, Enterococcus species and obligate anaerobes. 2,5 Methicillinresistant strains of S. aureus (MRSA) have previously been isolated mainly from hospitalized patients, but communityassociated cases are now common. 2 Moreover, strains with reduced susceptibility to vancomycin have been isolated in diabetic patients with foot infections. 6 Linezolid has been shown to have clinical utility in the treatment of diabetic patients with foot infections. 7 It possesses excellent in vitro activity against Gram-positive bacteria including MRSA and vancomycin-resistant enterococci, although most Gram-negative organisms are resistant to linezolid. 8 Furthermore, this agent has demonstrated good penetration into bone, joints, muscle and soft tissues in various patient populations. 9 -12 In this study, we further investigated the serum antibacterial activity and tissue penetration of linezolid in diabetic patients with severe foot infections requiring surgical intervention.

Subjects
Patients admitted to the hospital surgical service with severe diabetic foot infections were enrolled into this study. In addition, patients receiving antibiotic treatment with only intravenous linezolid (600 mg every 12 h) were eligible to participate in this trial. Each subject gave written informed consent that was approved by the hospital research review committee before entry into this investigation.

Serum and tissue samples
Blood samples were obtained on the day of surgery from each patient at 1, 3, 6, 9 and 12 h after the beginning of a 1 h intravenous infusion of linezolid. A blood sample was also obtained during surgery. Following centrifugation, serum samples were aliquotted and stored at 2708C until the time of analysis. The concentrations of linezolid in sera were measured by a validated HPLC assay (Dr Peloquin, National Jewish Medical and Research Center). 13 The serum curve for linezolid ranged from 0.5 to 30 mg/L. The percent coefficient of variation of a single standard concentration was 0.7%, and the overall validation precision across all standards was 1.0% to 4.4%.
A viable soft tissue sample was obtained from patients at their amputation site. These tissue specimens were wiped gently with dry gauze, placed into a pre-weighed vial and stored at 2708C until the time of analysis. For analysis, tissue samples were removed and sliced into smaller pieces (2-3 mm). They were then placed into two separate tubes and weighed. Samples were then treated with 0.5 mL of methanol/water (1:1) and vortex mixed for 30 s. Extracted tissue samples were treated with 0.5 mL of tissue lysis buffer (Buffer ATL; Qiagen, Hilden, Germany) vortexed for 30 s and then placed into a 508C hot water bath for 30 min. The concentration of linezolid from these samples was determined using a modification of the serum HPLC method. The recovery of linezolid from Buffer ATL-treated samples was found to be 96% to 103% from previous analyses.

Bacterial isolates
Well-defined isolates of MRSA were obtained from the National Institutes of Health through the NARSA (Network on Antimicrobial Resistance in S. aureus) programme. The MICs of linezolid and vancomycin were determined by microdilution methods according to the CLSI (formerly the NCCLS). 14 To ensure the stability of vancomycin-resistant strains, agar medium containing 4-6 mg/L of vancomycin was used for subculturing.
Inhibitory and bactericidal titres from sera collected from our patients were determined against these strains of MRSA according to CLSI methodology. 15 Serum was diluted with Mueller -Hinton broth and each determination was performed in duplicate. Wells with no visible growth and the first growth well were subcultured to supplemented Mueller -Hinton agar plates that were incubated for 2 days prior to counting colonies. Isolates were tested against serum collected at each time period for all subjects. The bacteriostatic titre-in-serum endpoint was determined as the highest dilution with no visible growth, and the bactericidal titre-in-serum endpoint was determined as the highest dilution of serum yielding 99.9% killing. The median and geometric mean bacteriostatic and bactericidal titres at each time period for these patients were calculated and used to determine serum inhibitory and bactericidal activity.

Pharmacokinetic analysis
An estimate of pharmacokinetic parameters from these serum samples was derived by non-compartmental analysis using the line of best fit for half-life and the linear trapezoidal rule for AUC 0 -12 (WinNon-Lin 4.0, Pharsight, Mountain View, CA, USA) based upon the assumption of steady-state and stable pharmacokinetics. 16

Results
Six patients (five men) with long-term diabetes were enrolled into this study. These subjects had an age range of 42 -81 years (mean, 61 years) and total body weights of 50-127 kg (mean, 98 kg). Each of these patients had peripheral vascular disease with diminished blood flow, documented from previous vascular studies and chronic foot infections requiring partial or complete amputation. Two patients had chronic renal insufficiency but none had liver disease. Serum glucose was controlled by insulin in each of these subjects during this study. All patients did well following surgery and were discharged from the hospital.
The pharmacokinetic parameters for linezolid in serum after multiple doses in these patients are given in Table 1. Tissue samples were obtained during surgery at 1 -9 h after the initiation of their linezolid infusion. Linezolid concentrations in these soft tissue specimens ranged from 18% to 78% (mean, 51%) of their simultaneous serum levels ( Table 2).
Sera from each patient were tested against five MRSA isolates with reduced susceptibility to vancomycin. The MIC for linezolid was 2 mg/L against all but one of these strains ( Table 3). The median serum inhibitory activities of linezolid were similar against each of these bacteria. Linezolid was found to exhibit rapid (1 h) and prolonged (12 h) inhibitory activity against each strain (Table 4). Median bactericidal activity was observed for at least 6 h (20/24 patients) against four of the five strains (Table 4). No bactericidal activity was observed against the MRSA VRS3 isolate.

Discussion
Linezolid is an oxazolidinone antimicrobial (MW 337) that is not highly protein-bound (30%) and has a moderate volume of distribution that approximates total body water. 17 Good penetration of linezolid into bone, fat, muscle, soft tissue and inflammatory blister fluid has been demonstrated in healthy volunteers and patients. 9 -12,18,19 Following multiple doses of linezolid, in vivo microdialysis studies found subcutaneous adipose tissue-to-plasma ratios of 0.9 in healthy subjects and critically ill patients. 12,18 In our patients, the serum pharmacokinetics of linezolid were similar to other investigations, but the mean tissue to serum ratio was only 0.51. 18,19 This finding was most likely due to impaired blood flow in our study population. These tissue samples were from the lower extremity in patients with peripheral vascular insufficiency at the infection (sample) site. Of more importance than the actual tissue concentration of an antibiotic is its antimicrobial activity at the site of infection. 4 In general, linezolid exhibits concentration-independent activity against Gram-positive cocci and the time interval of drug concentration that exceeds the MIC (t . MIC) and the area under the concentration -time curve to MIC (AUC/MIC) ratio are predictors of efficacy. 17,20 Data from animal models demonstrate that a t . MIC of 40% is a good prognostic indicator for a successful outcome. 17 Based upon linezolid's pharmacokinetic parameters after multiple doses, serum concentrations should remain above the MIC 90 (90% of strains) for Gram-positive bacteria for 40% of its dosing interval, even when these levels are diminished by 50% at the site of the infection. 17,21 These pharmacokinetic/pharmacodynamic characterizations have been indirectly supported in clinical trials. 22 For example, in diabetic patients with foot infections, the majority with foot ischaemia, linezolid exhibited good clinical cure rates including those patients with deep tissue and bone infections. 7 Infections caused by S. aureus isolates with reduced susceptibility to vancomycin are a growing problem and effective alternatives are needed. 23 Linezolid has good activity against staphylococci, but exhibits predominately bacteriostatic action in in vitro time -kill experiments. 24 In contrast to these observations, we found that serum can potentiate the antimicrobial effect of linezolid with resultant bactericidal activity. 16 In this current study, linezolid maintained median inhibitory activity for 12 h (17/30 samples) against each of the study isolates and median bactericidal activity for at least 6 h (50% of the dosing interval) for all but one of these S. aureus isolates. These findings are further supported from an in vitro pharmacodynamic model of    We believe that our findings have clinical relevance for several reasons. This ex vivo pharmacodynamic model integrates antimicrobial activity with pharmacokinetic parameters in patients. Furthermore, serum inhibitory and bactericidal titres allow for an evaluation of antibacterial activity in the presence of factors such as antibodies, complement, protein binding, as well as actual clinically relevant drug concentrations. In analysing serum bacteriostatic and bactericidal activity, patient sera are diluted serially in broth and incubated with the test strain of bacteria. 26 Moreover, these dilutions can simulate the concentration of drug at the site of infection. In general, our experiments found that linezolid exhibited median bacteriostatic and bactericidal activity in these patients at 1:4 and 1:2 dilutions, respectively, for at least 50% of the dosing interval against four of five MRSA strains tested ( Figure 1). These findings suggest that linezolid could be effective in the treatment of staphylococcal infections even when concentrations of this antibiotic are diminished due to impaired blood flow. It will also be important to monitor for adverse effects (e.g. thrombocytopenia and neuropathy) when using linezolid for extended treatment durations. 27