Correlation of MIC with Outcome for Candida Species Tested against Voriconazole: Analysis and Proposal for Interpretive Breakpoints

Abstract

Developing interpretive breakpoints for any given organism-drug combination requires integration of the MIC distribution, pharmacokinetic and pharmacodynamic parameters, and the relationship between the in vitro activity and outcome from both in vivo and clinical studies. Using data generated by standardized broth microdilution and disk diffusion test methods, the Antifungal Susceptibility Subcommittee of the Clinical and Laboratory Standards Institute has now proposed interpretive breakpoints for voriconazole and Candida species. The MIC distribution for voriconazole was determined using a collection of 8,702 clinical isolates. The overall MIC90 was 0.25 microg/ml and 99% of the isolates were inhibited at < or = 1 microg/ml of voriconazole. Similar results were obtained for 1,681 Candida isolates (16 species) from the phase III clinical trials. Analysis of the available data for 249 patients from six phase III voriconazole clinical trials demonstrated a statistically significant correlation (P = 0.021) between MIC and investigator end-of-treatment assessment of outcome. Consistent with parallel pharmacodynamic analyses, these data support the following MIC breakpoints for voriconazole and Candida species: susceptible (S), < or = 1 microg/ml; susceptible dose dependent (SDD), 2 microg/ml; and resistant (R), > or = 4 microg/ml. The corresponding disk test breakpoints are as follows: S, > or = 17 mm; SDD, 14 to 16 mm; and R, < or = 13 mm.