Relationship of MICs to efficacy of cefotaxime in treatment of Streptococcus pneumoniae infections

In June 1993, the National Committee for Clinical Laboratory Standards (NCCLS) recommended stringent new interpretive guidelines for antibiotics indicated for Streptococcus pneumoniae meningitis. To assess the predictive values of the recommended breakpoints, retrospective data were collected from patients who had S. pneumoniae infections and were treated with cefotaxime monotherapy. Susceptibilities based on the NCCLS interpretative categories were compared with clinical and bacteriologic outcomes. In 76 evaluable patients, the most common infections were bacteremia-septicemia (n = 49), meningitis (n = 37), and lower respiratory tract infection (n = 14). Under the NCCLS breakpoints proposed in 1993, 55 isolates would have been classed as susceptible to cefotaxime (MIC, or = 2 micrograms/ml). Of 75 cefotaxime-treated patients for whom cefotaxime MICs were recorded, 73 were clinically cured or improved (37 of 37 with meningitis and 36 of 38 with other infections). One case of bacteremia and one case of bone-and-joint infection were scored as therapeutic failures because initial monotherapy had to be modified because of an adverse drug reaction. Excluding these patients, there were 18 patients infected with S. pneumoniae that would have been classed as not fully susceptible (i.e., MICs > or = 0.5 microgram/ml); all of these patients were cured or improved. The results of this analysis demonstrate that successful treatment with cefotaxime did not correlate well with the guidelines for the susceptibility of pneumococcal isolates to either penicillin or cefotaxime established by the 1993 NCCLS breakpoint recommendations. Because of this study and other similar findings, the NCCLS adopted more clinically relevant guidelines in 1994.