Voriconazole concentrations and outcome of invasive fungal infections.

Twenty-five patients with proven or probable invasive fungal infections (IFIs) who experienced two or more episodes of voriconazole therapeutic drug monitoring (TDM) at a tertiary referral hospital were reviewed to explore the association between serum trough concentrations and outcomes of IFI. Microbiological and/or clinical success, in addition to IFI-related mortality, was assessed. We performed two separate analyses, one based on the initial trough voriconazole concentration at steady state, and the other on the median trough voriconazole concentration (derived from repeated TDM episodes) for each patient. Large interpatient and intrapatient variability of trough plasma voriconazole concentrations was observed, with no correlation between dose and concentration (r = 0.065). Classification and regression tree analysis revealed an association between IFI-related mortality and initial trough voriconazole concentrations, with patients more likely to die when their initial steady-state concentration was ≤0.35 mg/L (p 0.004; OR 11, 95% CI 2.9-41.2). Successful outcomes were more likely among patients with a median trough voriconazole concentration >2.2 mg/L (p 0.003; OR 2.7, 95% CI 1.4-5). Nineteen adverse events, with four severe events, were documented in 14 patients. Patients with severe adverse events had higher median voriconazole concentrations than the remaining cohort (2.38 mg/L vs. 1.30 mg/L; p <0.04). All adverse events resolved after cessation of voriconazole treatment. Our data suggest that voriconazole TDM is appropriate for all patients as soon as steady state is achieved. For non-responding patients with low trough concentrations, the association with IFI-related mortality indicates the need for dose adjustments to achieve and sustain voriconazole concentrations.