Application of different pharmacokinetic models to describe and predict pharmacokinetics of voriconazole in magellanic penguins following oral administration.

Aspergillosis is a condition causing serious morbidity and mortality in captive penguins and other bird species. It can be treated with antifungal drugs, such as voriconazole. However, the pharmacokinetics of voriconazole are variable between different animal and bird species. Therefore, the pharmacokinetics of voriconazole were investigated in this study in Magellanic penguins. Pharmacokinetic models were constructed and applied to predict the pharmacokinetics of voriconazole during long-term treatment in Magellanic penguins, since the voriconazole treatment duration in chronic aspergillosis cases can last up to several months. Plasma voriconazole concentration-time data from adult Magellanic penguins (Spheniscus magellanicus; n = 15) following a single oral (PO) dose of either 2.5 mg/kg or 5 mg/kg in a herring in three separate study periods 7-12 months apart were collected. Mean plasma voriconazole concentrations were above the targeted MIC for Aspergillus fumigatus for 2 hr following a single 2.5 mg/kg voriconazole dose while the plasma concentrations exceeded the MIC for least 24 hr following a 5 mg/kg dose. Nonlinear mixed-effects modeling was used to fit two pharmacokinetic models, one with first-order and another with saturable elimination, to the single-dose data. Fits were good for both, as long as dose was included as a covariate for the first-order model so that clearance was lower and the half-life longer for animals receiving the 5 mg/kg dose. Although the single-dose data suggested saturated elimination at higher concentrations, the model with saturable elimination did not predict plasma voriconazole concentrations well for a clinical aspergillosis case receiving long-term treatment, possibly because of induction of metabolizing enzymes with chronic exposure. Pharmacokinetic models should accurately predict plasma drug concentrations for different dosage regimens in order to be applicable in the field. Future studies should focus on determining clearance at steady-state to be able to refine the pharmacokinetic models presented here and improve model performance for long-term oral voriconazole administration in Magellanic penguins.

Pharmacokinetics of a single dose of voriconazole administered orally with and without food to red-tailed hawks (Buteo jamaicensus).

OBJECTIVE To determine the pharmacokinetics of voriconazole administered PO with or without food to red-tailed hawks (Buteo jamaicensus) and whether any observed variability could be explained by measured covariates to inform dose adjustments. ANIMALS 7 adult red-tailed hawks. PROCEDURES In a crossover study design, hawks were randomly assigned to first receive voriconazole (15 mg/kg, PO) injected into a dead mouse (n = 3; fed birds) or without food (4; unfed birds). Sixteen days later, treatments were reversed. Blood samples were collected at various points to measure plasma voriconazole concentrations by ultraperformance liquid chromatography. Pharmacokinetic data were analyzed by noncompartmental methods and fit to a compartmental model through nonlinear mixed-effects regression, with feeding status and body weight investigated as covariates. RESULTS Voriconazole was well absorbed, with quantifiable plasma concentrations up to 24 hours after administration. Mean plasma half-life was approximately 2 hours in fed and unfed birds. Administration of the voriconazole in food delayed absorption, resulting in a significant delay in time to maximum plasma concentration. The final compartmental model included a categorical covariate to account for this lag in absorption as well as body weight as a covariate of total body clearance (relative to unknown bioavailability). CONCLUSIONS AND CLINICAL RELEVANCE A single dose of voriconazole (15 mg/kg) administered PO to red-tailed hawks resulted in mean plasma voriconazole concentrations greater than the targeted value (1 µg/mL). Additional studies with larger sample sizes and multidose regimens are required before the model developed here can be applied in clinical settings.