Voriconazole pre-exposure selects for breakthrough mucormycosis in a mixed model of Aspergillus fumigatus-Rhizopus oryzae pulmonary infection

Mucormycosis is an uncommon fungal infection that has been increasingly reported in severely immunocompromised patients receiving Aspergillus-active antifungals. Although clinical studies and pre-clinical animal models have suggested a unique predisposition for breakthrough mucormycoses in patients receiving voriconazole, no study has specifically evaluated the selection dynamics of various Aspergillus-active antifungal classes in vivo. We utilized an Aspergillus fumigatus:Rhizopus oryzae (10:1) model of mixed fungal pneumonia in corticosteroid-immunosuppressed mice to compare the selection dynamics of daily liposomal-amphotericin B (L-AMB), micafungin (MCFG) and voriconazole (VRC) therapy. A. fumigatus and R. oryzae lung fungal burden were serially monitored in parallel using non-cross-amplifying quantitative real-time PC R assays for each fungal genus. Additionally, experiments were performed where the R. oryzae component of the mixed inoculum was serially passed on VRC-containing agar before animal infection. We found prior exposure to voriconazole in vitro, consistently resulted in a 1.5-2 log 10 increase in R. oryzae fungal burden by day +5 in vivo relative to animals infected with the non-VRC preexposed inoculum, irrespective of the antifungal-treatment administered in mice (p ≤ 0.02 all treatment groups). Mice infected with the VRC-preexposed inoculum and subsequently treated with saline or VRC had the highest mortality rates (82-86%), followed by MCFG (55%) then L-AMB (39%, p = 0.04 vs. control). However, in vivo treatment alone with voriconazole alone did not consistently increase the virulence of nonvoriconazole preexposed R. oryzae vs. controls. We conclude that exposure of R. oryzae sporangiospores to voriconazole in vitro modulates the subsequent growth rate and/or virulence of the fungus in vivo, which reduces effectiveness of Mucorales-active antifungals. The mechanisms underlying this phenotypic change are unknown