Background: The intravenous use of voriconazole requires coadministration with sulphobutylether-β-cyclodextrin, which may accumulate in patients with impaired renal function.
Methods: All adult patients treated with the same formulation of voriconazole for a minimum of 3 consecutive days were included. Renal function was assessed based on the creatinine level and the calculated creatinine clearance (CrCl). Change in renal function was calculated on days 3, 7, and end of treatment (EOT) and was defined based on the RIFLE criteria.
Results: Among 166 patients in whom baseline renal function was assessed, 42 (25.3%) had a CrCl <50 mL/min and received intravenous voriconazole, 77 (46.4%) had a CrCl ≥50 mL/min and received intravenous voriconazole, and 47 (28.3%) had a CrCl <50 mL/min and were treated with oral voriconazole. Renal function changed on days 3, 7, and EOT in 19 (11.4%), 14 (8.4%), and 28 (16.9%) patients, respectively. Multivariate analyses identified significant predictors of renal dysfunction: (1) day 3, hematologic malignancy (odds ratio [OR], 5.09, P = .01), fluconazole use within 30 days prior to voriconazole (OR, 6.21; P = .008), coadministration of penicillins (OR, 6.12; P = .03), and immunosuppressants (OR, 7.00; P = .002); (2) day 7, baseline liver impairment (OR, 5.30; P value = .004); (3) EOT, administration of penicillins (OR, 2.39; P = .04).
Conclusions: Voriconazole route of administration and baseline renal function were not predictors of worsening renal dysfunction on days 3, 7, and EOT. Underlying disease, baseline liver impairment, and concomitant administration of other drugs (eg, penicillins, fluoroquinolones, immunosuppressants) were the strongest predictors of renal dysfunction.