To obtain pharmacokinetic and pharmacodynamic data for vancomycin in a cohort of critically ill pediatric oncology patients, we analyzed 256 measurements of vancomycin concentrations in 94 patients. Variables were tested as possible risk factors for vancomycin-related nephrotoxicity or death for 28 days. We found the following: mean vancomycin trough serum concentration, 15.6 ± 12.4 μg/mL; mean vancomycin clearance, 0.16 ± 0.098 L/h/kg; and mean vancomycin distribution volume, 1.04 ± 0.11 L/kg. Only 13.6% of serum trough level measurements were between 15 and 20 μg/mL. The trough levels showed a strong correlation with the AUC (area under the curve of serum concentrations vs. time over 24 h to the minimum inhibitory concentration ratio), with a 94% positive predictive value for AUC/MIC ≥ 400, but only for MIC=1. The doses that are currently used (60 mg/kg/d) attained the therapeutic target (AUC/MIC ≥ 400) in only 56% of measurements, considering MIC=1. A serum trough level of ≥ 20 μg/mL was an independent risk for nephrotoxicity (P = 0.0008; odds ratio = 17.83). Vancomycin-related nephrotoxicity was a predictor of death for up to 28 days (P = 0.003, odds ratio = 7.68). Currently administered doses of vancomycin do not reach the therapeutic target for critical cancer patients, particularly if staphylococci isolates have a MIC>1.

Higher than usual vancomycin doses may be required to treat staphylococcal infections in children with oncologic/hematologic diseases. Since the best known predictor of efficacy is the AUC/MIC ratio, serum trough concentrations must be analyzed in conjunction with MICs of prevalent Staphylococci and pharmacokinetic tools such as Bayesian analysis.