Pharmacodynamics of Voriconazole for Invasive Pulmonary Scedosporiosis

Abstract

Scedosporium apiospermum is a medically important fungal pathogen that causes a wide range of infections in humans. There are relatively few antifungal agents that are active against Scedosporium spp. Little is known about the pharmacodynamics of voriconazole against Scedosporium Both static and dynamic in vitro models of invasive scedosporiosis were developed. Monoclonal antibodies that target a soluble cell wall antigen secreted by Scedosporium apiospermum were used to describe the pharmacodynamics of voriconazole. Mathematical pharmacokinetic-pharmacodynamic models were fitted to the data to estimate the drug exposure required to suppress the release of fungal antigen. The experimental results were bridged to humans using Monte Carlo simulation. All 3 strains of S. apiospermum tested invaded through the cellular bilayer of the in vitro models and liberated antigen. There was a concentration-dependent decline in the amount of antigen, with near maximal antifungal activity against all 3 strains being achieved with voriconazole at 10 mg/liter. Similarly, there was a drug exposure-dependent decline in the amount of circulating antigen in the dynamic model and complete suppression of antigen, with an area under the concentration-time curve (AUC) of approximately 80 mg · h/liter. A regression of the AUC/MIC versus the area under the antigen-time curve showed that a near maximal effect was obtained with an AUC/MIC of approximately 100. Monte Carlo simulation suggested that only isolates with an MIC of 0.5 mg/liter enabled pharmacodynamic targets to be achieved with a standard regimen of voriconazole. Isolates with higher MICs may need drug exposure targets higher than those currently recommended for other fungi.

Keywords: Scedosporium apiospermum; antifungal therapy; pharmacodynamics; pharmacokinetics; pneumonia; voriconazole.

FIG 1

Time course of antigen release in the in vitro static model. Data are means ± standard deviations for three inserts. The data were generated with strain 8353.

FIG 2

Concentration-response relationships for the three challenge strains of Scedosporium apiospermum. Data are means ± standard deviations. The broken and solid lines are the fit of an inhibitory sigmoid E_max model to the data from the alveolar and endothelial compartments, respectively. (A) Strain 8353; (B) strain CNM-CM6386; (C) strain CNM-CM6322.

FIG 3

Pharmacodynamics of voriconazole against Scedosporium apiospermum isolates from dynamic models of invasive pulmonary scedosporiosis. Time zero is the time of treatment initiation and was at 12 h postinoculation. Each line represents data from a single strain. Voriconazole was administered at time zero and every 12 h thereafter. The drug was administered as a bolus. Pharmacodynamic samples were obtained from the endothelial compartment of the model. The challenge strains were as follows: CNM-CM6386 (6386; red squares), CNM-CM6322 (6322; green triangles), and 8353 (blue circles). (A) Vehicle-treated control; (B) average AUC = 7.84 mg · h/liter, inducing a negligible effect; (C) average AUC = 15.62 mg · h/liter, inducing submaximal antifungal activity; (D) average AUC = 80.68 mg · h/liter, inducing nearly maximal antifungal activity.

FIG 4

Observed-predicted values from the mathematical models fitted to the pharmacokinetic and pharmacodynamic data from each strain. The broken line is the line of identity (where the observed concentration is equal to the predicted concentration).

FIG 5

(A) Relationship between voriconazole AUC_72–96 and the antifungal effect quantified in terms of the area under the antigen-time curve. The inhibitory sigmoid E_max model is given by the area under the antigen-time curve = 25.71 − {[16.69 · (AUC)^4.15]/[13.42^4.15 + (AUC)^4.15]} (r² = 0.86). (B) Relationship between voriconazole AUC_72–96/MIC and the antifungal effect, quantified in terms of the area under the antigen-time curve. The inhibitory sigmoid E_max model is given by the area under the antigen-time curve = 25.34 − {[17.14 · (AUC/MIC)^2.09]/[22.88^42.09 + (AUC/MIC)^2.09]} (r² = 0.82).

FIG 6

Probability of target attainment for a regimen of voriconazole at 6 mg/kg q12h i.v. for two dosages followed by 4 mg/kg q12h thereafter. The drug was infused over 1 h. A weight of 75 kg was used for the simulations, and the AUC_0–24 was determined at the end of day 5. The fraction of 5,000 simulated patients that achieved an AUC/MIC of >100 for each MIC value is shown by the solid squares. The MIC distribution for 77 S. apiospermum strains is shown by the open triangles. The overall expectation of target attainment for the population is 34%.