Journal of Antimicrobial Chemotherapy (2001) 48, 452-454
© 2001 The British Society for Antimicrobial Chemotherapy
Correspondence |
Susceptibility testing with linezolid by different methods, in relation to published ‘general breakpoints’
J Antimicrob Chemother 2001; 48: 452–454
Antibiotic Resistance Monitoring & Reference Laboratory, Central Public Health Laboratory, 61 Colindale Avenue, London NW9 5HT, UK
Sir,
Susceptibility testing with linezolid ought to be straightforward, since its MIC distributions are narrow, spanning only three or four doubling dilutions. Nevertheless, interpretative problems arise because several agencies have proposed breakpoints of 2 mg/L (Table 1
), whereas some workers find MICs of 4 mg/L for many staphylococci and enterococci, including even control strains.1,2 The potential for confusion is exacerbated, especially in surveys (where MICs are most likely to be determined), if a breakpoint of 2 mg/L is advocated but no method of susceptibility testing is specified (Table 1). The present study examines the effects of different testing methods on the linezolid MICs found for Gram-positive cocci, assessing whether particular conditions or media might bias the MICs, and lead to high rates of artefactual resistance.
|
The bacteria tested (n = 177) were chosen as recent clinical isolates and were collected from different hospitals so as to minimize the multiple inclusion of outbreak strains. They comprised: 42 Staphylococcus aureus (21 methicillin resistant, 21 susceptible), 42 coagulase-negative staphylococci (21 methicillin resistant, 21 susceptible), 21 Enterococcus faecalis (10 vancomycin resistant, 11 susceptible), 21 E. faecium (11 vancomycin resistant, 10 susceptible), 31 Streptococcus pneumoniae (10 penicillin resistant, 10 intermediate, 11 susceptible), 10 S. pyogenes, and 10 S. agalactiae. Various media were tested, with and without addition of lysed or whole equine or ovine blood (Table 2
), and with incubation in air or in 5% CO2. Agars were obtained from Oxoid (Basingstoke, UK), sheep blood from TCS Microbiology (Claydon, UK), horse blood from E&O Laboratories (Bonnybridge, Scotland, UK), linezolid from Pharmacia (Milton Keynes, UK), and Etests from Cambridge Diagnostics (Cambridge, UK). The objective was to test the effects of widely used sets of conditions, including those of the BSAC and NCCLS,3,4 not to assay the effect of every possible permutation of medium, blood and atmosphere.
|
Results are shown in Table 2
as MIC ranges, modes and geometric means. All the distributions, irrespective of the method, were narrow, spanning four dilutions or fewer, and were unimodal and unskewed. No frank resistance (MIC 
8 mg/L) was seen in any organism. Variations in the observed MICs in relation to the media and conditions were small, such that all the mode and geometric mean values obtained by the different methods were always within two doubling dilutions of each other. The most notable trend was for the MICs by Etest to be one doubling dilution lower than by agar and broth dilution, regardless of whether Etests were used on Mueller–Hinton or IsoSensitest agars. This seemingly reflected the fact that the agar and broth MICs were read to the point of final termination of growth whereas, following the manufacturer's advice, the linezolid Etests were read to the point where 80% of the growth was inhibited (difficult to judge precisely!). Had the same reading policy been adopted for agar and broth MICs, these too would have been reduced by approximately one dilution.
Several recorded MIC ranges reached 4 mg/L and a modal MIC of 4 mg/L was recorded for S. aureus as tested by the NCCLS broth method. In previous studies using the BSAC method we have commonly found MICs of 4 mg/L for enterococci. A situation whereby MICs may exceed a 2 mg/L breakpoint (Table 1) contingent on the precise methodology or upon the single-dilution run-to-run variation conventionally accepted in susceptibility testing evidently is unsatisfactory both for the treatment of patients and the surveillance of resistance. Breakpoints of 4 mg/L, as now adopted by the BSAC and EUCAST, seem more appropriate, although caution may be warranted if an MIC of 4 mg/L is recorded by a laboratory that mostly finds MICs of 1 mg/L for the same species. A breakpoint of 4 mg/L is also supported by pharmacodynamic data, which indicate that a 600 mg dose of linezolid achieves a serum drug concentration >4 mg/L throughout the 12 h inter-dose interval. Finally, it should added that linezolid MICs for resistant E. faecium selected in therapy5 are reported to be 32–64 mg/L, well above any of the breakpoints detailed in Table 1.
Acknowledgements
We are grateful to Pharmacia Corp., Milton Keynes, for financial support.
Notes
* Corresponding author. Tel: +44-20-8200-4400; Fax: +44-20-8358-3292; E-mail: DLivermore{at}phls.nhs.uk 
References
1
.
Johnson, A. P., Warner, M. & Livermore, D. M. (2000). Activity of linezolid, a novel oxazolidonone, against multi-resistant gram-positive bacteria. Journal of Antimicrobial Chemotherapy 45, 225–30.
2 . Zurenko, G. E., Yagi, B. H., Schaadt, R. D., Allison, J. W., Kilburn, J. O., Glickman, S. E. et al. (2000). In vitro activities of U-100592 and U-100766, novel oxazolidinone antibacterial agents. Antimicrobial Agents and Chemotherapy 40, 839–45.[Abstract]
3
.
British Society for Antimicrobial Chemotherapy Working Party. (1991). A guide to sensitivity testing. Journal of Antimicrobial Chemotherapy 27, Suppl. D, 1–50.
4 . National Committee for Clinical Laboratory Standards. (2000). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically—Fifth Edition: Approved Standard M7-A5. NCCLS, Wayne, PA.
5 . Gonzales, R. D., Schreckenberger, P. C., Graham, M. B., Kelkar, S., DenBesten, K. & Quinn, J. P. (2001). Infections due to vancomycin-resistant Enterococcus faecium resistant to linezolid. Lancet 357, 1179.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. M. Boak, J. Li, C. R. Rayner, and R. L. Nation Pharmacokinetic/Pharmacodynamic Factors Influencing Emergence of Resistance to Linezolid in an In Vitro Model Antimicrob. Agents Chemother., April 1, 2007; 51(4): 1287 - 1292. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. R. Wilson, J. A. Cepeda, S. Hayman, T. Whitehouse, M. Singer, and G. Bellingan In vitro susceptibility of Gram-positive pathogens to linezolid and teicoplanin and effect on outcome in critically ill patients J. Antimicrob. Chemother., August 1, 2006; 58(2): 470 - 473. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Buerger, N. Plock, P. Dehghanyar, C. Joukhadar, and C. Kloft Pharmacokinetics of unbound linezolid in plasma and tissue interstitium of critically ill patients after multiple dosing using microdialysis. Antimicrob. Agents Chemother., July 1, 2006; 50(7): 2455 - 2463. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Draghi, D. J. Sheehan, P. Hogan, and D. F. Sahm In Vitro Activity of Linezolid against Key Gram-Positive Organisms Isolated in the United States: Results of the LEADER 2004 Surveillance Program Antimicrob. Agents Chemother., December 1, 2005; 49(12): 5024 - 5032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Honeybourne, C. Tobin, G. Jevons, J. Andrews, and R. Wise Intrapulmonary penetration of linezolid J. Antimicrob. Chemother., June 1, 2003; 51(6): 1431 - 1434. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||