Experience with Voriconazole Therapeutic Drug Monitoring During the 2012 Fungal Meningitis Outbreak

Background: In 2012, nationwide distribution of contaminated methylprednisolone steroid injections resulted in an outbreak of over 750 associated fungal infections, primarily due to Exserohilum rostratum. Interim guidelines for empiric treatment recommending both high-dose intravenous voriconazole (VCZ) and liposomal amphotericin B were provided by the CDC. Due to the non-linear pharmacokinetics and saturable clearance of VCZ, therapeutic drug monitoring (TDM) is recommended and has been associated with improvements in both efficacy and safety. The aim of this study is to examine the relationship between VCZ pharmacokinetics and clinical outcomes specific to patients receiving extended courses of therapy for CNS infections.

Methods: This was a retrospective, single-center cohort study of all patients with previous exposure to ≥ 1 contaminated steroid injection and treated at Carilion Clinic. Classification and regression tree (CART) analysis was used to identify variables associated with patient response and adverse event occurrence. Variables such as type of infection, Charlson Comorbidity Index score, age, and associated VCZ level were incorporated into the interaction analysis.

Results: A total of 267 serum and CSF VCZ levels were obtained for 34 patients over a mean treatment duration of 102.5 ± 11.1 days. Twenty-three patients (67%) met criteria for complete treatment response on VCZ. However, no association between VCZ level and treatment response was found by CART analysis. Similarly, VCZ levels were not associated with adverse event occurrence. There was, however, an association between duration of therapy and side effect onset. CNS toxicity and hepatotoxicity typically developed within the first week of treatment (<5.5 days) while alopecia and periostitis only occurred with prolonged (>51.5 days) VCZ exposure. The most common side effects observed in this study were CNS toxicity (e.g. hallucinations or confusion), hepatotoxicity, alopecia, vision changes, and rash.

Conclusion: VCZ concentrations were not associated with treatment outcomes or adverse drug effects in this real-world study of immunocompetent patients with CNS fungal infections. These results call into question the applicability of current VCZ TDM practices to long treatment courses in the outpatient setting.