Antimicrobial resistance in nosocomial isolates of Staphylococcus haemolyticus.

Staphylococcus haemolyticus is frequently cultured from hospitalized patients and is characterized by resistance to multiple antimicrobial agents. We found that S. haemolyticus represented 70 of 524 (13%) coagulase-negative staphylococcal isolates identified by the clinical microbiology laboratories of two hospitals over 2 months. S. haemolyticus isolates were recovered from wounds (44%), urine (26%), blood (10%), and other sources (20%). All S. haemolyticus isolates were tested for susceptibility to six antimicrobial agents; 77% were resistant to three or more agents, and 41% were resistant to five or six agents. In addition, among 47 multiply resistant isolates, high MICs (greater than or equal to 6.25 micrograms/ml) of vancomycin (62% of isolates) and teicoplanin (91% of isolates) were found. DNA probes which were derived from S. epidermidis or S. aureus and which contained sequences associated with resistance to antimicrobial agents were used to detect specific genes in the total cellular and plasmid DNAs of 10 resistant S. haemolyticus isolates. Resistance gene probes and the numbers of resistant isolates hybridizing were as follows: methicillin, 10 of 10; gentamicin, 9 of 10; erythromycin, 7 of 10; and trimethoprim, 0 of 10. Genes for resistance to methicillin were found only in chromosomal locations, genes for resistance to gentamicin were found in both chromosomal and plasmid locations, and genes for resistance to erythromycin were found in plasmid locations only. With the exception of trimethoprim resistance determinants, similar genes were found among concurrently isolated multiply resistant S. epidermidis isolates from our hospitals. S. haemolyticus is a potentially important nosocomial species which readily acquires antimicrobial resistance genes and which shares, to some extent, in a common gene pool with S. epidermidis.

Molecular epidemiology of trimethoprim resistance among coagulase-negative staphylococci.

A 42% (70 of 167 isolates) incidence of resistance to 20 micrograms of trimethoprim per ml was found among clinical isolates of coagulase-negative staphylococci from two hospitals. A specific trimethoprim resistance gene probe from a conjugative Staphylococcus aereus plasmid was used to investigate the location of the trimethoprim resistance gene among 29 isolates. In 14 trimethoprim-resistant isolates, the probe hybridized with only chromosomal DNA, in 9 it hybridized with only plasmid DNA, and in 1 isolate both plasmid and chromosomal sequences showed hybridization. In five isolates there was no hybridization of the probe with either chromosomal or plasmid DNA. Four of these five nonhybridizing isolates were Staphylococcus haemolyticus. In contrast, all 22 Staphylococcus epidermidis isolates tested hybridized with the probe. The presence of the trimethoprim resistance gene in a chromosomal location was correlated with a lower MIC (median, 80 micrograms/ml) than when it was plasmid encoded (median, 1,250 micrograms/ml). Restriction endonuclease mapping as well as DNA hybridization of cloned plasmid and chromosomal DNA showed that there were 2.7 kilobases of common DNA in the two loci. This included the 500 base pairs of DNA mediating trimethoprim resistance and a total of 2.2 kilobases of 3'- and 5'-flanking sequences. The presence of the same gene and flanking sequences in chromosomal and plasmid locations suggests that the trimethoprim resistance determinant is translocated among different genetic loci.

Teicoplanin compared with vancomycin for treatment of experimental endocarditis due to methicillin-resistant Staphylococcus epidermidis.

This study compared teicoplanin with vancomycin without and with gentamicin and/or rifampin for treatment of experimental endocarditis due to methicillin-resistant Staphylococcus epidermidis. In rabbits treated for three days and killed 12 hr after the last doses of antimicrobial agents, no significant difference in reducing bacterial titers of vegetations was detected between vancomycin and teicoplanin without and with gentamicin and/or rifampin. Addition of gentamicin and/or rifampin to vancomycin or teicoplanin significantly reduced bacterial titers of vegetations compared with vancomycin or teicoplanin alone. Addition of rifampin alone or gentamicin plus rifampin was significantly more effective than addition of gentamicin alone. In rabbits treated for three days and killed seven days after the last doses of antimicrobial agents, no significant difference in sterilizing vegetations was detected between vancomycin and teicoplanin with gentamicin and/or rifampin. However, there was a trend (probably due to the longer elimination half-life of teicoplanin in serum) that clearly favored teicoplanin over vancomycin. Teicoplanin plus rifampin without or with gentamicin is at least as effective as vancomycin plus rifampin without or with gentamicin for treatment of experimental endocarditis due to methicillin-resistant S. epidermidis.