Influence of testing methodology on the tigecycline activity profile against presumably tigecycline-non-susceptible Acinetobacter spp.

OBJECTIVES: To compare the tigecycline activity profile against Acinetobacter spp. by Etest versus broth microdilution in isolates with high Etest MIC. METHODS: Acinetobacter spp. isolates with tigecycline MICs of >or=0.5 mg/L determined by commercially developed Etests strips (January 2006 to July 2007) in five Spanish hospitals were considered. Values were rounded to the nearest upper double-dilution. Susceptibility by broth microdilution following CLSI (formerly NCCLS) recommendations, as the reference method, was determined in a central laboratory. BSAC breakpoints were used: susceptible 2 mg/L. RESULTS: One hundred and forty-eight isolates were collected: 12 isolates with a tigecycline Etest MIC of 0.5 mg/L, 14 with 1 mg/L, 86 with 2 mg/L, 31 with 4 mg/L and 5 with 8 mg/L. Isolates with Etest MICs of 0.5-1 mg/L showed the same values by broth microdilution. Among isolates with Etest MICs of 2 mg/L, only 5.8% of strains showed the same value by both methods (88.4% showed values that were one or two dilutions lower by microdilution). None of the 36 isolates with Etest MICs of 4-8 mg/L showed the same value by both methods, with values at least two dilutions lower by microdilution. Weak correlation (R = 0.238; P or=2 mg/L for Acinetobacter spp. since strains with Etest MICs of 2-4 mg/L are susceptible when tested by microdilution. False non-susceptibility by Etest may exclude tigecycline as a therapeutic option in a field where multiresistance is the rule.