GLYCOPEPTIDE TOLERANCE IN STAPHYLOCOCCUS-AUREUS

Treatment failures with vancomycin prompted us to investigate the phen omenon of tolerance to glycopeptides in recent clinical isolates of St aphylococcus aureus. We used both MBC/MIC determinations and time-kill measurements to study tolerance to vancomycin and teicoplanin in 35 b lood or heart valve isolates of S. aureus from patients with endocardi tis or bacteraemia. There was generally good agreement between vancomy cin tolerance indicated by an MBC:MIC ratio of greater than or equal t o 32 and by less than or equal to 90% kill after 6 h incubation in the presence of 20 mg/L vancomycin. However, two isolates were tolerant a ccording to their MBC:MIC ratios but nontolerant as judged by time-kil l measurements. Seven of 15 methicillin-resistant S. aureus (MRSA) iso lates but only two of 20 methicillin-susceptible ones were tolerant as judged by time-kill experiments (chi(2) = 4.27 With Yates' correction , P = 0.04). Seven of the 16 isolates from patients with endocarditis were tolerant, compared with only two of the 19 isolates from patients with other conditions (chi(2) = 3.43 with Yates' correction, P = 0.06 ). Within the endocarditis and non-endocarditis subgroups, tolerance w as associated more frequently with methicillin resistance than with su sceptibility, but the numbers were too small for the differences to be statistically significant. Most of the vancomycin-tolerant isolates w ere also tolerant to teicoplanin. We conclude that glycopeptide tolera nce is a real phenomenon in S. aureus, particularly amongst MRSA isola tes, and can be reliably determined by our method of time-kill analysi s. Tolerance may compromise glycopeptide therapy of serious S. aureus infection and should be taken into account when deciding treatment.