Relationship between the level of acquired resistance to gentamicin and synergism with amoxicillin in Enterococcus faecalis

In enterococci, intrinsic low-level resistance to gentamicin does not abolish synergism with a cell wall-active antibiotic while high-level resistance due to acquired aminoglycoside-modifying enzymes does. To study the impact of intermediate levels of resistance to gentamicin (64 < MIC < 500 mu g/ml), we selected in vitro three consecutive generations of mutants of Enterococcus faecalis JH2-2 with MICs of gentamicin at 128 mu g/ml for G1-1477, 256 mu g/ml for G2-1573, and 512 mu g/ml for G3-1688. E. faecalis 102, which is highly resistant to gentamicin by enzymatic inactivation was used as control. In in vitro killing curves experiments, gentamicin concentrations allowing bactericidal activity and synergism in combination with amoxicillin increased from 4 mu g/ml (1/16th the MIC), 16 mu g/ml (one-eighth the MIC), 64 mu g/ml (one-quarter the MIC), and 256 mu g/ml (one-half the MIC) for strains JH2-2, G1-1477, G2-1573 and G3-1688, respectively. As expected, no bactericidal effect of the combination or synergism could be obtained with strain 102. In rabbits with aortic endocarditis caused by strain G1-1477 or G2-1573, combination therapy with amoxicillin and gentamicin was significantly more active than amoxicillin alone (P < 0.05) but not in those infected with the strains G3-1688 and 102. Thus, intermediate levels of resistance to gentamicin was not associated with a loss of a beneficial effect of the gentamicin-amoxicillin combination in vivo even though higher concentrations of gentamicin were necessary to achieve in vitro synergism. Therefore, the use of an MIC of 500 mu g/ml as a clinical cutoff limit to predict in vivo benefit of the combination remains a simple and effective tool.