JAC Advance Access originally published online on January 12, 2006
Journal of Antimicrobial Chemotherapy 2006 57(3):498-505; doi:10.1093/jac/dki489
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Epidemiological MIC cut-off values for tigecycline calculated from Etest MIC values using normalized resistance interpretation
1 Clinical Microbiology—MTC, Karolinska Institute, Karolinska Hospital L2:02, 17176 Stockholm, Sweden; 2 Clinical Microbiology, Malmö University Hospital, Malmö, Sweden; 3 Department of Infectious Diseases, Karolinska Hospital, Karolinska Institute, SE-17176 Stockholm, Sweden; 4 Division of Clinical Microbiology, Department of Molecular and Clinical Medicine, Faculty of Health Sciences, SE-58185 Linköping, Sweden
* Corresponding author. Tel: +46-8-51774910; Fax: +46-8-308099; E-mail: goran.kronvall{at}ki.se
Received 24 September 2005; returned 9 November 2005; revised 14 December 2005; accepted 20 December 2005
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Objectives: To apply the normalized resistance interpretation (NRI) method to Etest MIC results which have higher precision than conventional log2 dilution MIC tests due to the inclusion of intermediate values. If successful, NRI might provide an objective tool for the definition of epidemiological MIC cut-off values.
Methods: MICs of tigecycline and other antimicrobial agents were determined for 4771 clinical isolates comprising five Gram-positive and 13 Gram-negative species or species groups using the Etest. Histograms of MIC values were constructed for each species and NRI calculations were applied to them. An upper MIC limit of 2.5 SD above the theoretical mean of the normalized distribution was used for setting the epidemiological cut-off values.
Results: Calculated cut-off values for wild-type strains were between 0.11 and 0.96 mg/L for Gram-positive species, and between 0.44 and 8.3 mg/L for Gram-negative species, except for Pseudomonas aeruginosa, which had a cut-off value of 450 mg/L, consistent with earlier reports on the lack of activity of tigecycline against this species.
Conclusions: NRI offers an objective method for the analysis of MICs produced using Etests and the determination of epidemiological MIC cut-off values.
Keywords: antimicrobial susceptibility tests , MIC breakpoints , drug resistance , clinical microbiology , wild-type MIC distributions
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Interpretive breakpoints for susceptibility reporting by clinical microbiology laboratories were previously set for an antimicrobial agent with no consideration of bacterial species differences. In recent years such differences have been appreciated and species-related interpretive breakpoints are issued more frequently.1,2 A rational consequence of such thinking is to limit interpretation as susceptible to wild-type populations, a concept which originally made use of the term ‘microbiological breakpoints’3 but which later used the amended term ‘epidemiological cut-off values’ in a move towards harmonization of European standards.4,5 Isolates of a given bacterial species which differ from the wild-type population might be considered resistant to the drug in question but could sometimes be designated ‘intermediately susceptible’ or even ‘susceptible’, if a therapeutic effect was evident in clinical trials.
When trying to define the wild-type population of a given bacterial species, there is a problem as to how to differentiate susceptible from resistant isolates. With regard to zone diameter distributions of disc diffusion tests, a new method, normalized resistance interpretation (NRI), holds promise.6–8 However, when NRI is applied to doubling dilution MIC distributions, the relatively few observations on the susceptible side of the distribution peak compared with disc test results preclude any accurate NRI calculations due to the limited number of points on the curve (G. Kronvall, unpublished data). A diffusion-based MIC method which includes 1/2log2 values, the Etest, was used in in vitro susceptibility studies of tigecycline in Sweden. We applied NRI on results from these studies.
Tigecycline1 (GAR-936) is a new semisynthetic derivative of tetracycline and the first in the class of glycylcycline antibiotics. Tigecycline is a broad-spectrum antibiotic with activity against the majority of Gram-positive and Gram-negative pathogenic bacteria.9–16 It has demonstrated significant bactericidal activity against Haemophilus influenzae, Escherichia coli and Staphylococcus aureus. Its activity against Pseudomonas aeruginosa, however, is poor with an MIC50 of 8 mg/L11 or 16 mg/L.12 Tigecycline is active against microorganisms which carry the classical tetracycline resistance genes, both tetracycline efflux and ribosomal protection determinants.17 It is active against penicillin-resistant Streptococcus pneumoniae and vancomycin-resistant Enterococcus spp. The application of NRI for the determination of epidemiological MIC cut-off values in the present studies was of interest, both for testing the NRI method using Etest results, and for the determination of epidemiological tigecycline cut-off values for different bacterial species to compare with the conventional setting of interpretive breakpoints.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Bacterial isolates
Isolates of five Gram-positive and 13 Gram-negative pathogenic bacterial species were collected during a 1 year period (November 2003–October 2004) at the microbiology laboratories of three university hospitals in Sweden (Karolinska University Hospital, Stockholm, Linköping University Hospital, Linköping, and Malmö University Hospital, Malmö). Up to 100 isolates of each species were collected from general wards and less from the ICUs. Duplicate isolates from patients were excluded. Species identification followed regular procedures in the accredited microbiology laboratories.18
Antimicrobial susceptibility tests
The antimicrobial susceptibility of the isolates was determined using Etests (AB Biodisk, Solna, Sweden) for panels of antimicrobials according to species. All isolates were tested for tigecycline susceptibility. Large (150 mm) agar plates with Mueller–Hinton medium were inoculated with suspensions of the strains and Etest strips were applied. After incubation the Etests were read according to the manufacturer's instructions. Values in-between the regular log2 dilution series were recorded as such without approximation upwards.
Interpretive MIC breakpoints for tigecycline were S 
0.5 mg/L for S. aureus; S 0.25 mg/L for Streptococcus spp. other than S. pneumoniae; S 0.25 mg/L for Enterococcus species, and S 2 mg/L and R 
8 mg/L for Enterobacteriaceae (tigecycline approved for injection by the US FDA on 16 June 2005. Breakpoints by the FDA, USA, acc. to tigecycline package insert). For S. pneumoniae the MIC breakpoint used for comparative purposes in the present investigation was S 0.25 mg/L, for coagulase-negative staphylococci (CoNS) it was S 0.5 mg/L, and for Gram-negative species other than Enterobacteriaceae, S 2 mg/L and R 8 mg/L.
MIC50 and MIC90 values were calculated using MIC values adjusted upwards according to the regular log2 scale for comparative purposes. The IC50 and IC90 values (intrapolated concentration for inhibiting 50% and 90% of the isolates, respectively) were also determined in order to provide a more precise calculation of the inhibitory activity of the drug.19,20 The results were analysed species-wise as were the distribution plots of MIC values.
NRI
The ideal MIC distribution of wild-type isolates was reconstructed using NRI (Bioscand AB, Täby, Sweden, International Patent Application WO 02/083935 A1). NRI utilizes the low MIC side of the susceptible strains in a distribution histogram of MIC values for a mathematical reconstruction of the ideal wild-type peak.6–8 The method was first developed for inhibition zone diameter histograms but the same principles apply also to MIC distributions. Etests have a higher precision than regular doubling dilution MIC determinations and might therefore be better suited for NRI calculations. For the procedure, the MIC values are first converted to zone-equivalents (ZEs) using the simple procedure: MIC 
log2 MIC plus 10 doubled subtract this value from 50. This gives a theoretical zone diameter equivalent for the purpose of calculation with low values for high MICs and can be directly entered into the NRI computer program for zone diameters and the regular NRI calculations are performed. The wild-type population is thereby defined and a limit for susceptibility can then be set as a number of standard deviations (SDs) from the mean. Parametric statistical measures, such as the mean and SD, have been shown earlier to describe these histogram populations accurately.21 A limit of 2 SD above the mean will theoretically include 97.7% of the susceptible isolates, 2.5 SD will include 99.4% and 3 SD will include 99.9%. The 2.5 SD limit was chosen for setting cut-offs in the present studies.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
MIC determinations and susceptibility interpretations of clinical isolates
In all, 4771 isolates were collected, of which 1915 were Gram-positive and 2856 were Gram-negative (Table 1). The MIC50 and MIC90 values of tigecycline for the Gram-positive isolates, which belonged to five species, were <0.5 mg/L. Three species showed 100% susceptibility and two species showed 96.4% (CoNS) and 99.7% (S. pneumoniae) susceptibility. The relatively high degree of homogeneity among Gram-positive isolates is illustrated in Figure 1.
|
|
The MIC values for the Gram-negative isolates were generally higher and showed wider variation (Figure 2). As shown in Figure 2(d), P. aeruginosa in particular constituted a population for which MICs were relatively high compared with the majority of the Gram-negative species (MIC50 16 mg/L, MIC90 32 mg/L). MIC50 and MIC90 values for the other Gram-negative species ranged from 0.25 mg/L for E. coli to 1.0 and 2.0–4.0 mg/L, respectively, for Proteus mirabilis, Proteus vulgaris and Morganella morganii. The proportion of Gram-negative isolates susceptible to tigecycline ranged from 100% (E. coli, Klebsiella oxytoca, Citrobacter spp., H. influenzae) to 86.8% (M. morganii), except for P. aeruginosa where only 5% of isolates were susceptible using MIC breakpoints for comparative purposes. The frequent notations of intermediate susceptibility for Gram-negative isolates indicated that the MIC breakpoints were probably close to the MIC distribution of the actual populations of the strains. This was evident from the MIC distributions for the different Gram-negative species.
|
Analysis of MIC distributions of Gram-positive species using NRI
The tigecycline MIC distributions of Gram-positive isolates are shown in Figure 1 with the higher resolution offered by the Etest. The homogeneity of the species populations in the low MIC range reflects the virgin status of clinical isolates to this new antimicrobial. The mean values of the true MIC values calculated using logarithmic distributions were 0.086, 0.16, 0.084, 0.067 and 0.041 mg/L, for S. aureus, CoNS, Enterococcus faecalis, Enterococcus faecium and S. pneumoniae, respectively (Table 1). When NRI calculations were applied to the distributions, the Etest values proved sufficient for the mathematics to work (Figure 1, normalized distributions shown as line plots). The number of value pairs in the critical regression between probit values and converted MIC values was 4, 5, 4, 4 and 3, for the five species, respectively, which gave correlation coefficients (r values) of 0.959, 0.992, 0.993, 0.992 and 0.992, respectively, for the regressions. The NRI-calculated means of the theoretical normal distributions of susceptible isolates were 0.11, 0.15, 0.074, 0.075 and 0.038 mg/L, respectively (Table 1). These values were very close to the real distributions reflecting the homogeneity of these susceptible populations. There was only one isolate which was clearly outside, an S. pneumoniae isolate for which the MIC value was 0.5 mg/L.
An upper MIC limit for the susceptible populations of the five Gram-positive species was calculated using an upper 2.5 SD limit of the NRI-derived normal distributions. The values obtained were 0.5, 1.0, 0.5, 0.5 and 0.125 mg/L, for the five species, respectively (Table 1). Our calculations also suggest an MIC cut-off value for S. pneumoniae of S 0.25 mg/L. CoNS do not comprise a single species and hence the distribution shows a wider spectrum of MIC values. Still, the coefficient of variation (CV) of the SD over the mean (calculated using the converted MIC values) was only 6%, slightly higher than the CV for the other Gram-positive species, which were between 3% and 4%. With a margin for this slight heterogeneity the MIC cut-off value could be set to S 2.0 mg/L.
Analysis of MIC distributions of Gram-negative species using NRI
The NRI-generated normal distributions of the 2855 Gram-negative isolates belonging to 13 different bacterial species were plotted as line graphs on the MIC histogram distributions and representative examples are shown in Figure 2. The mathematics of the calculations worked well using the Etest-generated MIC values and the numbers of value pairs were between 3 and 6, except for P. aeruginosa where there were nine value pairs used. The coefficients of variation were likewise between 3.2% and 7.2%, except for P. aeruginosa which gave a CV of 17.2%. This species showed a wide distribution of MIC values with a trailing also on the more susceptible side from 0.38 mg/L to the true mean of 12.9 or the NRI mean of 17.8 mg/L (Figure 2d and Table 1). These means are far above the interpretive MIC breakpoints for susceptibility among Enterobacteriaceae and the whole population of P. aeruginosa isolates therefore might be considered resistant to tigecycline.
The NRI-generated normal distributions of the Gram-negative species agreed well with the MIC histogram distributions (Figure 2). The true MIC means and the NRI-generated normal distribution means for all species except P. aeruginosa were all below the breakpoint for the Gram-negative organisms used in these studies (S 2.0 mg/L), namely between 0.16 (E. coli) and 1.27 mg/L (P. mirabilis) and between 0.15 (E. coli and Acinetobacter species) and 1.36 mg/L (Serratia marcescens), respectively. If all these homogeneous populations of wild-type isolates are shown to respond to tigecycline treatment, then the 2.5 SD upper limits calculated from the NRI, ranging between 0.44 and 5.4 mg/L for most of the species and 8.3 mg/L for S. marcescens, might serve as guidelines for the setting of proper epidemiological cut-off values (Table 1).
When the nine species belonging to the Enterobacteriaceae were analysed using the epidemiological cut-off values from NRI analysis (Table 1) the proportion of susceptible isolates rose to 100% for seven species and to 99.3% and 99.7% for E. coli and Enterobacter spp., respectively. These figures are more in line with wild-type populations, but for use in susceptibility testing they also require clinical trials to have shown tigecycline to be effective against such isolates. They also support the suspicion raised above that the tentative breakpoints issued by the FDA might give a certain rate of false resistance for Enterobacteriaceae species. The need for and importance of clinical trials holds particularly for Proteus species, M. morganii and S. marcescens where the MIC values are close to the FDA breakpoints. The NRI results for S. marcescens gave a broad peak with 6 value points and the highest CV among Enterobacteriaceae and the 2 SD limit was as high as 8.34 mg/L. It should be noted that no isolate of S. marcescens had a tigecycline MIC > 3 mg/L.
Populations of both S. marcescens and P. aeruginosa isolates showed a marked trailing towards lower MIC values. Such differences among presumably wild-type isolates could be due to differences in, for instance, active efflux pumps, which might also affect the susceptibility levels to other antimicrobial agents. A comparison between the MIC levels of tigecycline and the other antimicrobials used for testing P. aeruginosa was therefore made. In these comparisons the MIC values were transformed to log2 + 9 values, as regularly used in regression studies. Tigecycline was compared with piperacillin/tazobactam, ceftazidime, imipenem, meropenem, gentamicin and ciprofloxacin and the correlation coefficients obtained were 0.831, 0.802, 0.699, 0.608, 0.894 and 0.596, respectively. Gentamicin showed the strongest correlation, although not a very strong one. The distribution of MIC values for S. marcescens and P. aeruginosa should be analysed further.
NRI analysis of MIC distributions among populations with isolates resistant to other antibiotics
The application of NRI to MIC distributions was thus mathematically possible using Etest-generated tigecycline MIC values according to the results of testing both Gram-positive and Gram-negative isolates. It is expected that NRI generates a theoretical distribution reflecting the susceptible wild-type population only, something that couldn't be critically tested with tigecycline where most isolates were fully susceptible. Other Etest-generated antimicrobial MIC distributions with resistant isolates were therefore analysed and a typical example is shown in Figure 3 with S. pneumoniae isolates tested against tetracycline. The 349 isolates showed a tetracycline susceptibility level of 91.7% with 8.3% resistance. The NRI-generated distribution gave an upper 2.5 SD value of 0.56 mg/L which would mean an S-limit of 1 mg/L, compared with the regular MIC breakpoint S 2.0, R > 2.0 mg/L for pneumococci (SRGA, www.srga.org). The NRI generated MIC-limit would include 91.4% as tetracycline susceptible. Thus, NRI correctly identified the susceptible population also when resistant isolates were present in the population.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The main purpose of the present investigation was to evaluate the possibility that MIC values determined using the Etest with its more precise measurements might constitute a basis for applying the NRI method, in order to provide an objective determination of epidemiological MIC cut-off values. NRI has previously been applied successfully to species populations with antimicrobial susceptibility results in the form of inhibition zone diameter values.6–8 Previous attempts at using regular log2 dilution MIC values for NRI calculations have not been successful (G. Kronvall, unpublished data). The results of the present studies show that Etest-based NRI calculations do work mathematically and NRI might therefore be used also for analysis of antimicrobial susceptibility using MIC data. It is particularly promising for testing new antimicrobials, like tigecycline, where interpretive breakpoints often are unavailable at an early stage. Epidemiological cut-off values calculated using an objective method like NRI could be useful in early studies where the identification of isolates different from the wild-type population is of interest. It should be pointed out, however, that NRI calculations of epidemiological cut-off values might also be of great value when analysing susceptibility to older antimicrobials using the vast source of disc diffusion test results available in clinical microbiology laboratories world-wide. Such laboratory-specific and species-related cut-off values for susceptibility are independent of the methodology used and can therefore provide comparable results.
The present investigation provides tigecycline susceptibility data for bacterial pathogens isolated in Sweden. When compared with studies of tigecycline susceptibility of isolates from other parts of the world, the MIC50 and MIC90 values obtained are very similar.9–12,15,16 The only consistent deviations were for P. mirabilis, P. vulgaris and M. morganii, with MIC50/MIC90 values in our studies of 1/4, 1/2 and 1/4 mg/L, respectively, but 4/8, 4/4–8 and 1–4/4–8 mg/L in other studies.9–12 It is not clear from the other studies whether the isolates of these species constituted homogeneous, wild-type populations, or if the presence of isolates with resistance mechanisms raise the MIC50 and MIC90 values. A clinical isolate of P. mirabilis with a tigecycline MIC of 4 mg/L showed increased expression of an acrRAB gene fragment resulting in increased efflux.22 In another study two clinical isolates of M. morganii with tigecycline MIC values of 4 mg/L were shown to overexpress acrA resulting in increased efflux.23 Whether differences in efflux pump expression might explain the slightly higher susceptibility in Swedish isolates remains to be determined. In such a case the present calculation of epidemiological MIC cut-off values for tigecycline and these three species might not be representative of data from other countries. The fact that Etest results for some occasional species–antimicrobial combinations show consistent deviations up or down compared with reference MIC methods might also be operating in these cases. However, no studies so far seem to indicate such a trend.
The US FDA has provided MIC breakpoints for interpretation of susceptibility for three of the five Gram-positive species and when these values were compared with the NRI-calculated MIC cut-off values for susceptibility, the S. aureus breakpoint was identical, whereas the Enterococcus cut-off values were one step higher in our calculations compared with the MIC breakpoint of the FDA (Table 1). The similarity in mean values between the three species, both for the isolates and for the NRI distributions, supports a similar susceptibility MIC breakpoint for these species.
The tigecycline MIC histograms exemplified in Figures 1 and 2 illustrate the fact that wild-type populations have a restricted range of MIC values in any given species and also that the locations of these populations differ between different bacterial species. As a consequence, the NRI-generated 2.5 SD upper limits for susceptibility differ between different species (Table 1). A tendency in guidelines given by reference authorities in recent years, for example by CLSI (formerly NCCLS) and SRGA (Swedish Reference Group for Antibiotics, http://www.srga.org/), is to relate interpretive zone or MIC breakpoints to individual species. SRGA introduced species-related zone breakpoints in the late 1980s,1,24 a feature that was introduced in Sweden, in the late 1970s.25 The problem of defining a susceptible (usually) wild-type distribution of MIC values for isolates of a given species has not been solved earlier in a rational way. The wild-type distributions of antimicrobial MIC values for different species and drugs given by EUCAST represent a successful effort to make such distributions available, but the methods involved are not declared (http://www.escmid.org/). NRI is a mathematical, automatic procedure which means that it will produce objective results devoid of subjective influences. Its capacity to define the theoretical, normalized distribution of wild-type isolates in an MIC histogram based on Etest results offers a new tool for the definition of epidemiological MIC cut-off values.
Statistical methods have recently been applied to define wild-type populations using log2 antimicrobial MIC distributions of isolates species-wise, with promising results.26 This is an iterative method where the most resistant values are deleted step-wise and the remaining population analysed regarding a Gaussian distribution. When some particular parameters indicate a good fit, then the remaining distribution is considered the wild-type population of isolates for that particular combination of species and antimicrobial. This means that also regular doubling dilution MIC values are possible to analyse. One possible draw-back would be the difficulties which slightly resistant variants will pose producing a more or less visible shoulder on the distribution. The NRI method, on the other hand, is a purely mathematical reconstruction in one step and has earlier shown a great potential in analyzing zone diameter histograms. It is independent of the presence of even slightly resistant isolates, but requires that the tests are performed with the very best quality. In the present studies we have shown that it can also be applied to MIC distributions using Etest values. The two methods will therefore complement each other keeping in mind their differences in power.
Transparency declarations
Other than the research grant from Wyeth International, Sweden, supporting the present studies, there are no economic or other links between this company and the participating scientists.
|
Acknowledgements |
|---|
This work was supported by a research grant from Wyeth International, Sweden.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
1. Ringertz S, Olsson-Liljequist B, Kahlmeter G et al. Antimicrobial susceptibility testing in Sweden. II. Species-related zone diameter breakpoints to avoid interpretive errors and guard against unrecognized evolution of resistance. Scand J Inf Dis Suppl 1997; 105: 8–12.[Medline]
2. National Committee for Clinical Laboratory Standards. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically—Fifth Edition. Approved Standard, M7-A5. Wayne, PA: NCCLS, 2000.
3. Williams JD. Prospects for standardisation of methods and guidelines for disc susceptibility testing. Eur J Clin Microbiol Inf Dis 1990; 9: 496–501.
4.
Kahlmeter G, Brown DF, Goldstein FW et al. European harmonization of MIC breakpoints for antimicrobial susceptibility testing of bacteria. J Antimicrob Chemother 2003; 52: 145–8.
5. Kahlmeter G, Brown DFJ. Harmonization of antimicrobial breakpoints in Europe—can it be achieved? Clin Microbiol Newsl 2004; 26: 187–92.
6. Kronvall G, Kahlmeter G, Myhre E et al. A new method for normalized interpretation of antimicrobial resistance from disk test results for comparative purposes. Clin Microbiol Infect 2003; 9: 120–32.[CrossRef][Web of Science][Medline]
7. Kronvall G. Determination of the real standard distribution of susceptible strains in zone histograms. Int J Antimicrobial Agents 2003; 22:7–13.[CrossRef][Web of Science][Medline]
8. Joneberg J, Rylander M, Galas MF et al. Analysis of parameters and validation of method for normalized interpretation of antimicrobial resistance. Int J Antimicrobial Agents 2003; 21: 525–35.[CrossRef][Web of Science][Medline]
9. Zhanel GG, Homenuik K, Nichol K et al. The Glycylcyclines. A Comparative Review with the Tetracyclines. Drugs 2004; 64: 63–88.[CrossRef][Web of Science][Medline]
10.
Petersen PJ, Jacobus NV, Weiss WJ et al. In vitro and in vivo antibacterial activities of a novel glycylcycline, the 9-t-butylglycylamido derivative of minocycline (GAR-936). Antimicrob Agents Chemother 1999; 43: 738–44.
11. Gales AC and Jones RN. Antimicrobial activity and spectrum of the new glycylcycline, GAR-936 tested against 1,203 recent clinical bacterial isolates. Diagn Microbiol Infect Dis 2000; 36: 19–36.[CrossRef][Web of Science][Medline]
12.
Milatovic D, Schmitz FJ, Verhoef J et al. Activities of the glycylcycline tigecycline (GAR-936) against 1,924 recent European clinical bacterial isolates. Antimicrob Agents Chemother 2003; 47: 400–4.
13. Deshpande LM, Gales AC, Jones RN. GAR-936 (9-t-butylglycylamido-minocycline) susceptibility test development for streptococci, Haemophilus influenzae and Neisseria gonorrhoeae: preliminary guidelines and interpretive criteria. Int J Antimicrobial Agents 2001; 18: 29–35.[Web of Science][Medline]
14.
Betriu C, Culebras E, Rodriguez-Avial I et al. In vitro activities of tigecycline against erythromycin-resistant Streptococcus pyogenes and Streptococcus agalactiae: mechanisms of macrolide and tetracycline resistance. Antimicrob Agents Chemother 2004; 48: 323–5.
15. Fritsche TR, Kirby JT, and Jones RN. In vitro activity of tigecycline (GAR-936) tested against 11,859 recent clinical isolates associated with community-acquired respiratory tract and gram-positive cutaneous infections. Diagn Microbiol Infect Dis 2004; 49: 201–9.[CrossRef][Web of Science][Medline]
16. Bouchillon SK, Hoban DJ, Johnson BM et al. In vitro evaluation of tigecycline and comparative agents in 3049 clinical isolates: 2001 to 2002. Diagn Microbiol Infect Dis 2005; 51: 291–5.[CrossRef][Web of Science][Medline]
17.
Bauer G, Berens C, Projan SJ et al. Comparison of tetracycline and tigecycline binding to ribosomes mapped by dimethylsulphate and drug-directed Fe2+ cleavage of 16S rRNA. J Antimicrob Chemother 2004; 53: 592–9.
18. Kronvall G, Hagelberg A. Numerical evaluation of minimal biochemical test combinations for the identification of Enterobacteriaceae species. APMIS 2002; 110: 451–7.[CrossRef][Web of Science][Medline]
19. Reyn A, Weis Bentzon M, Ericsson H. Comparative investigations of the sensitivity of N. gonorrhoeae to penicillin. Acta Pathol Microbiol Scand 1963; 235–55.
20.
Soothill JS, Ward R, Girling AJ. The IC50: an exactly defined measure of antibiotic sensitivity. J Antimicrob Chemother 1992; 29: 137–9.
21. Kronvall G, Ringertz S, Nyström S et al. Comparison of 30 µg and 120 µg gentamicin disks for the prediction of gentamicin resistance in Enterococcus faecalis. APMIS 1991; 99: 887–92.[Web of Science][Medline]
22.
Visalli MA, Murphy E, Projan SJ et al. AcrAB multidrug efflux pump is associated with reduced levels of susceptibility to tigecycline (GAR-936) in Proteus mirabilis. Antimicrob Agents Chemother 2003; 47: 665–9.
23.
Ruzin A, Keeney D, Bradford PA. AcrAB efflux pump plays a role in decreased susceptibility to tigecycline in Morganella morganii. Antimicrob Agents Chemother 2005; 49: 791–3.
24. Olsson-Liljequist B, Forsgren A. Antimicrobial susceptibility testing in Sweden. I. The work of the Swedish Reference Group for Antibiotics (SRGA and SRGA-M). Scand J Infect Dis Suppl 1997; 105: 5–7.[Medline]
25. Kronvall G, Runehagen A. Förbättrad känslighetsbedömning av bakterier medelst artrelaterat SIR-system (Improved determination of antibiotic susceptibility of bacteria using species-related SIR interpretation). Läkartidningen 1981; 78: 3483–5.[Medline]
26. Turnidge J, Kahlmeter G, Kronvall G. Statistical characterisation of bacterial wild-type distributions of MIC-values and the determination of epidemiological cut-off values. Clin Microbiol Infect 2006; 12: in press.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Kelesidis, D. E. Karageorgopoulos, I. Kelesidis, and M. E. Falagas Tigecycline for the treatment of multidrug-resistant Enterobacteriaceae: a systematic review of the evidence from microbiological and clinical studies J. Antimicrob. Chemother., November 1, 2008; 62(5): 895 - 904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Turnidge and D. L. Paterson Setting and Revising Antibacterial Susceptibility Breakpoints Clin. Microbiol. Rev., July 1, 2007; 20(3): 391 - 408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Turnidge and G. Bordash Statistical Methods for Establishing Quality Control Ranges for Antibacterial Agents in Clinical and Laboratory Standards Institute Susceptibility Testing Antimicrob. Agents Chemother., July 1, 2007; 51(7): 2483 - 2488. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. N. Jones, M. J. Ferraro, L. B. Reller, P. C. Schreckenberger, J. M. Swenson, and H. S. Sader Multicenter Studies of Tigecycline Disk Diffusion Susceptibility Results for Acinetobacter spp. J. Clin. Microbiol., January 1, 2007; 45(1): 227 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Vahaboglu, F. Budak, M. Kasap, G. Gacar, S. Torol, A. Karadenizli, F. Kolayli, and C. Eroglu High prevalence of OXA-51-type class D {beta}-lactamases among ceftazidime-resistant clinical isolates of Acinetobacter spp.: co-existence with OXA-58 in multiple centres J. Antimicrob. Chemother., September 1, 2006; 58(3): 537 - 542. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||