Voriconazole-induced inhibition of the fungicidal activity of amphotericin B in Candida strains with reduced susceptibility to voriconazole: an effect not predicted by the MIC value alone

Abstract

An antagonistic effect of voriconazole on the fungicidal activity of sequential doses of amphotericin B has previously been demonstrated in Candida albicans strains susceptible to voriconazole. Because treatment failure and the need to switch to other antifungals are expected to occur more often in infections that are caused by resistant strains, it was of interest to study whether the antagonistic effect was still seen in Candida strains with reduced susceptibility to voriconazole. With the hypothesis that antagonism will not occur in voriconazole-resistant strains, C. albicans strains with characterized mechanisms of resistance against voriconazole, as well as Candida glabrata and Candida krusei strains with differences in their degrees of susceptibility to voriconazole were exposed to voriconazole or amphotericin B alone, to both drugs simultaneously, or to voriconazole followed by amphotericin B in an in vitro kinetic model. Amphotericin B administered alone or simultaneously with voriconazole resulted in fungicidal activity. When amphotericin B was administered after voriconazole, its activity was reduced (median reduction, 61%; range, 9 to 94%). Levels of voriconazole-dependent inhibition of amphotericin B activity differed significantly among the strains but were not correlated with the MIC values (correlation coefficient, -0.19; P = 0.65). Inhibition was found in C. albicans strains with increases in CDR1 and CDR2 expression but not in the strain with an increase in MDR1 expression. In summary, decreased susceptibility to voriconazole does not abolish voriconazole-dependent inhibition of the fungicidal activity of amphotericin B in voriconazole-resistant Candida strains. The degree of interaction could not be predicted by the MIC value alone.

FIG. 1.

Expression levels of drug resistance genes in C. albicans strain CA4_MIC256 compared with those in the azole-susceptible strain CAF2-1. 28S and 18S RNA is shown as the control for identical RNA loading.

FIG. 2.

Time-kill plots in the in vitro kinetic model of voriconazole (VRC) and amphotericin B (AMB) regimens against C. albicans strains CA1_MIC8 (A), CA2_MIC8 (B), CA3_MIC2 (C), CA4_MIC256 (D), and CA5_MIC0.004 (E); C. glabrata strains CG1_MIC0.5 (F) and CG2_MIC4 (G); and C. krusei strains CK1_MIC0.5 (H) and CK2_MIC4 (I). Colony count values below the limit of detection were set to 0.5 log₁₀ CFU/ml.

FIG. 2.

Time-kill plots in the in vitro kinetic model of voriconazole (VRC) and amphotericin B (AMB) regimens against C. albicans strains CA1_MIC8 (A), CA2_MIC8 (B), CA3_MIC2 (C), CA4_MIC256 (D), and CA5_MIC0.004 (E); C. glabrata strains CG1_MIC0.5 (F) and CG2_MIC4 (G); and C. krusei strains CK1_MIC0.5 (H) and CK2_MIC4 (I). Colony count values below the limit of detection were set to 0.5 log₁₀ CFU/ml.

FIG. 3.

Time-kill plots of C. albicans CA1_MIC8 (A), CA2_MIC8 (B), CA3_MIC2 (C), CA4_MIC256 (D), and CA5_MIC0.004 (E) exposed to 2.0 mg/liter of amphotericin B (AMB) after preexposure to 5.0 mg/liter voriconazole for 24 h. Time zero is the time point for AMB exposure. Colony count values below the limit of detection were set to 0.5 log₁₀ CFU/ml.