Streptococcus pneumoniae bacteraemia: pharmacodynamic correlations with outcome and macrolide resistance--a controlled study.

There are few data on macrolide pharmacodynamics in pneumococcal infections. We evaluated pneumococcal area under the inhibitory concentration-time curve (AUIC) values at the point of hospital admission in 59 bacteraemic patients failing in the community and in 98 bacteraemic controls without macrolide exposure. The area under the 24-h concentration-time curve (AUC24) was calculated for each patient using age, weight and daily dose; using minimum inhibitory concentrations (MICs), the values of AUIC (i.e. AUC24/MIC) were then computed. Clinical and outcome information was also collected in hospital. Five of six patients who died of pneumococcal bacteraemia in hospital received azithromycin, with a mean AUIC of 8.1 prior to hospital admission. Resistant isolates were recovered in 35 (59%) macrolide failures and in only 28 (29%) controls (P=0.001). Azithromycin AUICs averaged 10 in failure patients and 17 in controls. For clarithromycin and erythromycin, the mean AUIC values in failures were 31 and 53, respectively, and the AUIC in controls was >100. Low AUIC values against Streptococcus pneumoniae precede macrolide failures in the community. Patient factors do not predict these outcomes and thus the most likely explanation for macrolide failure in the community is inadequate macrolide activity in patients who receive these antibiotics for treatment of organisms that are not sufficiently susceptible.