Should serum voriconazole (Vori) levels be monitored in allogeneic hematopoietic stem cell transplant (HSCT) recipients?

Abstract

Vori has significant activity against Candida and Aspergillus. It is metabolized by the hepatic cytochrome P450 enzymes 2C19, 2C9 and 3A4. CYP2C19, which plays a significant role in the metabolism, exhibits significant genetic polymorphism. As a result, some patients metabolize Vori poorly. The clinical significance of this and resultant possible increases in Vori levels is unknown, and the utility of monitoring Vori levels is unclear. We monitored steady-state trough serum Vori levels in 22 allogeneic HSCT recipients using a new HPLC assay (Pennick et al. Antimicrob Agents Chemother 2003;47:2348–50). All had received non-myeloablative conditioning with 100 mg/m2 melphalan (+50 mg/kg cyclophosphamide if not prior autograft). GVHD prophylaxis comprised cyclosporine/tacrolimus with mycophenolate mofetil. Vori was started at 200 mg BID PO in 20 patients, and 400 mg BID for a day followed by 200 mg BID in 2. One had aspergillosis, and the rest were receiving Vori prophylactically or empirically. Patients had drug levels checked once (n = 12), twice (n = 8), or ≥3 times (n = 2); 5 days to 9 months (median 10 days) after starting Vori or dose modification. The intention was to achieve a level greater than 0.5–1.0 μg/mL corresponding to the in vitro MIC90 of Vori for Aspergillus spp. The 22 baseline levels ranged from 0.2 to 6.8 μg/mL (median 1.1). Overall, the 36 levels were 0.2–6.8 (median 1.7); with 5 levels <0.5. An increase in the dose to 300 mg twice daily in 3 patients (2 with levels <0.5 and 1 with a level of 1.0 in the presence of aspergillosis) increased the level 6- to 8-fold. There was no significant correlation between Vori levels and serum creatinine, bilirubin, and ALT. Vori levels correlated with alkaline phosphatase (r = 0.44; P = 0.008) and AST (r = 0.52; P = 0.001). Since liver dysfunction is relatively common after allogeneic HSCT, it was not possible to determine if elevated AST and alkaline phosphatase levels were the cause or the consequence of higher Vori levels. Based on these data, we conclude that trough Vori levels vary considerably between patients with a small proportion having potentially subtherapeutic levels on standard doses. We suggest checking serum Vori levels in patients receiving the drug for confirmed fungal infections. Additionally, patients developing elevation of AST or alkaline phosphatase on therapy should also have Vori levels monitored. Vori levels should also be checked if the drug dose is changed.