Broth Microdilution Susceptibility Testing of Francisella tularensis: Quality Control Limits for Nine Antimicrobial Agents and Three Standard Quality Control Strains

Standardization of susceptibility testing methods for potential agents of bioterrorism such as Francisella tularensis is necessary for therapeutic guidance in the event of an outbreak with a potentially resistant isolate (4). The low incidence of naturally occurring cases coupled with the hazardous nature of the organism has precluded the development of susceptibility tests for F. tularensis (3). Due to the fastidious nutritional requirements of this organism, the medium must be enriched with l-cystine (1, 5, 6). Additives such as Isovitalex (Becton-Dickinson, Sparks, Md.), which contains l-cystine, are used in media to satisfy such in vitro growth prerequisites. The purpose of this study was to propose quality control (QC) ranges for nine antimicrobial agents diluted in cation-adjusted Mueller-Hinton broth (CAMHB)-2% IsoVitalex with three National Committee of Clinical Laboratory Standards (NCCLS)-recommended QC strains under incubation conditions conducive to the growth of F. tularensis.

MIC testing was performed according to the recommendations of the NCCLS (7, 8). Broth microdilution trays were commercially prepared by TREK Diagnostic Systems (Cleveland, Ohio) to contain serial dilutions of multiple drugs diluted in each of three different lots of CAMHB containing 2% IsoVitalex (vitamin B₁₂ [0.01g], l-glutamine [10.0 g], adenine [1.0 g], guanine hydrochloride [0.03 g], p-aminobenzoic acid [0.013 g], NAD [0.25 g], thiamine pyrophosphate [0.1 g], ferric nitrate [0.02 g], thiamine hydrochloride [0.003 g], l-cysteine hydrochloride [25.9 g], l-cystine [1.1 g], dextrose [100.0 g], purified water [1 liter]). Aseptic adjustment of the pH to a range of 7.3 ± 0.1 was required after the addition of the IsoVitalex. The antimicrobial agents are listed in Table 1, Table 2, and Table 3. The trays were then frozen and shipped to the eight participating laboratory sites which are identified in the acknowledgments. On separate days of testing, each of three QC strains, Staphylococcus aureus ATCC 29213, Escherichia coli ATCC 25922, and Pseudomonas aeruginosa ATCC 27853, was inoculated into the MIC trays. The MIC trays were incubated at 35°C in ambient air and read visually at 24 h and again at 48 h. The laboratory at each study site tested the three organisms in three separate lots of CAMHB for 10 consecutive days. During the study, laboratories performed colony counts to insure proper inoculation concentrations. The median colony counts were 4.4 × 10⁵ (range, 3.1 × 10⁵ to 6.8 × 10⁵) for S. aureus ATCC 29213, 3.4 × 10⁵ (range, 2.4 × 10⁵ to 6.3 × 10⁵) for E. coli ATCC 25922, and 4.6 × 10⁵ (range, 3.0 × 10⁵ to 6.9 × 10⁵) for P. aeruginosa ATCC 27853.

Tables 1 to 3 represent the frequency distributions for the MICs of the antimicrobial agents tested with each of the three QC strains. No significant variability was observed among results obtained using the three different lots of CAMHB. MICs read at 24 and 48 h differed by no more than one log₂ dilution interval. Three dilution ranges were proposed whenever there was a unimodal distribution of the values, and a 4-dilution range was proposed whenever there was a bimodal distribution of results (2, 8). With one exception, the MICs for S. aureus attained >96% distribution within the proposed ranges. The exception was nalidixic acid: a log₂ dilution shift in mode after 24 and 48 h of incubation resulted in MICs at the limit of or beyond the testing ranges (Table 1). For the E. coli control strain, >97% of MICs were within the proposed ranges. For P. aeruginosa, >97% of MICs were within the proposed ranges although no ranges were recommended for chloramphenicol or nalidixic acid due to off-scale results. No ranges were proposed for trimethoprim-sulfamethoxazole because of excessive interlaboratory variability. None of these agents are among those considered to comprise the first line of defense against tularemia.

On the basis of the data provided by the eight laboratories, the Subcommittee on Antimicrobial Susceptibility Testing of the NCCLS approved the QC ranges listed in Table 4. For MIC results for F. tularensis read after 24 h of incubation, the 24-h QC ranges should be used; for results read after 48 h, only the 48-h QC ranges should be used.

The first step in standardization of the MIC assay for F. tularensis is the establishment of a broth medium. It is then possible to select MIC ranges for commonly employed QC strains when that medium and the incubation environment required for testing clinical isolates are used. Because F. tularensis requires l-cystine for growth, the CAMHB used for other species for conventional MIC testing methods (1, 4, 5) requires modification. Defined supplements such as Isovitalex provide the required l-cystine, and the NCCLS has approved the use of such a broth formulation for susceptibility testing of F. tularensis.

The choice of antimicrobial agents in this study focused on the reported activities of these agents in the treatment of tularemia. Although streptomycin, gentamicin, chloramphenicol, and tetracycline were the preferred therapeutic agents in the past, newer antimicrobial agents with greater bactericidal and intracellular activity, including newer fluoroquinolones, macrolides, and doxycycline, are now recommended (5, 9).

• Copyright © 2004 American Society for Microbiology