Comparative pulsed-field gel electrophoresis typing of gentamicin-resistant and -susceptible methicillin-resistantStaphylococcus aureus strains isolated in France between 1991 and 1998. Changes in antibiotic susceptibility

Abstract

Using macrorestriction of genomic DNA and pulsed-field gel electrophoresis, we examined 504 non-redundant, infection-causing human isolates of methicillin-resistant Staphylococcus aureus susceptible (GS: 238 isolates) or resistant to gentamicin (GR: 266 isolates). The strains were isolated at Albert Chenevier Hospital (Créteil, France) between 1 January 1991 and 31 December 1998. Their susceptibility to erythromycin, lincomycin, tetracycline, rifampicin, fusidic acid and fosfomycin was also studied. Seventy-six genotypes were identified (percentage similarity<80). Ten types, each containing at least eight strains, predominated. GR strains showed higher genetic polymorphism than GS strains: the 266 GR isolates belonged to 67 genotypes, five of which predominated (44, 42, 38, 30 and 15 isolates); the 238 GS isolates belonged to only 18 types, four of which predominated (112, 83, 11 and 10 isolates). Fifty-six percent of GR strains (34 Gt) were resistant to erythromycin, lincomycin, tetracycline and rifampicin, and were isolated at relatively stable frequencies. Resistance to five antibiotics studied (susceptible to fusidic acid) was observed among 16.5% of GR strains. The frequency of strains with this profile diminished from 30% in the early 1990s to 10% in 1998. One hundred and twenty-six GS isolates were susceptible to all six antibiotics; this profile was only found from 1993 onwards, and was increasingly frequent (60% of GS strains in 1996). Resistance to erythromycin and lincomycin only was found in 70 GS isolates; this profile accounted for approximately half the isolates in 1992/1993 and only one-third in 1998. These results, obtained over an eight-year period, show an overall increase in antibiotic susceptibility. They confirm the spread of two major clones of MRSA-GS and support the hypotheses that GS strains derive from GR strains that have lost the aac6′-aph2″ gene; and that GS strains are genetically related to those that were present before the use of gentamicin and persisted at a low frequency until 1992–1993.