Voriconazole concentration is inversely correlated with corticosteroid usage in immunocompromised patients.

BACKGROUND: Voriconazole (VRCZ) is a broad-spectrum antifungal agent that can be administered both orally and intravenously. A high level of intra- and interindividual variability in serum drug therapeutic concentrations has been reported. We analyzed the influence of corticosteroid use on serum VRCZ concentration by assessing the correlation between corticosteroid dose and VRCZ trough level.

METHODS: Immunocompromised patients treated with VRCZ with or without corticosteroid use from June 2010 to March 2017 (6 years and 8 months) were reviewed in our institute. VRCZ and the corticosteroids were administered orally or intravenously. Corticosteroid treatment was considered as "concurrent" only if it was administered within 3 days before or after the initiation of VRCZ. All corticosteroids were converted to a prednisolone-based dosage according to their anti-inflammatory effect (relative glucocorticoid activity). VRCZ concentration was measured at the trough point on the day of steady state.

RESULTS: A total of 85 samples were obtained from 38 patients. The medical records of 23 women and 15 men, with a median age of 61 years (range: 35-86), were reviewed. Twenty-one patients (55.3%, 21/38) received chemotherapy, and 28 (73.7%, 28/38) received corticosteroid therapy. The average and median daily doses of corticosteroids were 59.2 and 89.8 mg, respectively (range: 0.714-377). VRCZ concentration was inversely correlated with corticosteroid dose (r = -.26). The VRCZ concentration was significantly decreased in the corticosteroid user compared to nonuser (P = .013). We evaluated the association of VRCZ concentration and corticosteroid dose in 3 patients among our cohorts for whom more than 5 points of data were available. Intraindividual analysis revealed the trough level of VRCZ concentration decreased as the corticosteroid dose increased. The patients' underlying diseases were hematological diseases (n = 25) and immunological diseases (n = 13). Among accrual patients, no patients were undergoing stem cell transplantation, and no patients were treated with known confounders such as calcineurin inhibitors or phenytoin.

CONCLUSIONS: VRCZ might be metabolized by an enzyme induced by corticosteroid treatment; therefore, intraindividual variation in VRCZ metabolism should be considered prior to concurrent treatment, especially for patients with hematological malignancies treated with corticosteroids.