Pharmacokinetics and pharmacodynamics of fluconazole for cryptococcal meningoencephalitis: implications for antifungal therapy and in vitro susceptibility breakpoints

Abstract

Fluconazole is frequently the only antifungal agent that is available for induction therapy for cryptococcal meningitis. There is relatively little understanding of the pharmacokinetics and pharmacodynamics (PK-PD) of fluconazole in this setting. PK-PD relationships were estimated with 4 clinical isolates of Cryptococcus neoformans. MICs were determined using Clinical and Laboratory Standards Institute (CLSI) methodology. A nonimmunosuppressed murine model of cryptococcal meningitis was used. Mice received two different doses of fluconazole (125 mg/kg of body weight/day and 250 mg/kg of body weight/day) orally for 9 days; a control group of mice was not given fluconazole. Fluconazole concentrations in plasma and in the cerebrum were determined using high-performance liquid chromatography (HPLC). The cryptococcal density in the brain was estimated using quantitative cultures. A mathematical model was fitted to the PK-PD data. The experimental results were extrapolated to humans (bridging study). The PK were linear. A dose-dependent decline in fungal burden was observed, with near-maximal activity evident with dosages of 250 mg/kg/day. The MIC was important for understanding the exposure-response relationships. The mean AUC/MIC ratio associated with stasis was 389. The results of the bridging study suggested that only 66.7% of patients receiving 1,200 mg/kg would achieve or exceed an AUC/MIC ratio of 389. The potential breakpoints for fluconazole against Cryptococcus neoformans follow: susceptible, ≤ 2 mg/liter; resistant, >2 mg/liter. Fluconazole may be an inferior agent for induction therapy because many patients cannot achieve the pharmacodynamic target. Clinical breakpoints are likely to be significantly lower than epidemiological cutoff values. The MIC may guide the appropriate use of fluconazole. If fluconazole is the only option for induction therapy, then the highest possible dose should be used.

Fig 1

Model predictions for the pharmacokinetics of fluconazole (FZL) in plasma (solid line) and in the cerebrum (broken line) in mice receiving 125 mg/kg/day (A) and 250 mg/kg/day (B). Raw data from individual mice are shown by the open circles (plasma) and open squares (cerebrum).

Fig 2

Pharmacodynamics of fluconazole against four strains of Cryptococcus neoformans. Mice received no fluconazole (controls) or received 125 mg/kg and 250 mg/kg/day of fluconazole orally. In all cases, the horizontal and vertical axes represent time and the cerebral fungal burden, respectively. The solid line is the fit of the respective mathematical models to the data from each strain. Raw data (open circles) are means ± standard deviations (error bars) for 3 mice. With the exception of strain F/13186, approximately 250 mg/kg/day is required to achieve stasis (i.e., the fungal burden at the start of therapy is the same as the end of therapy).

Fig 3

Relationship between the fluconazole area under the concentration-time curve at steady state (AUC_0–24) and decline in fungal burden in the brain relative to controls. Data are means ± standard deviations (error bars). The solid line is the fit of the sigmoid Emax model. The experimental data are placed in a clinical context by showing the means ± standard deviations of AUC values for 5,000 simulated patients receiving 200, 400, 600, and 800 mg of fluconazole/day.

Fig 4

Relationship between the ratio of the fluconazole area under the concentration-time curve (AUC) to MIC (AUC/MIC ratio) and the decline in fungal density in the brain. Data are means ± standard deviations (error bars). The solid line is the fit of the sigmoid Emax model. The experimental data are placed in a clinical context by showing the means ± standard deviations of the AUC/MIC ratios for 5,000 simulated patients receiving 1,200 mg of fluconazole/day and infected with strains with various MIC values given at the top of the figure. As the MIC increases from 2 to 32 mg/liter, there is a progressive decline in the AUC/MIC ratio and a smaller corresponding antifungal effect.

Fig 5

Attainment of the pharmacodynamic target (AUC/MIC ratio of ≥389) as a function of MIC. The solid squares delineate the MIC distribution of Cryptococcus neoformans estimated using CLSI methodology. The solid circles show the proportion of 5,000 simulated patients receiving 1,200 mg of fluconazole/day that have an AUC/MIC ratio of ≥389.3.