The Gastrointestinal Tract Is a Major Source of Echinocandin Drug Resistance in a Candida glabrata Colonization Mouse Model

BACKGROUND: Gastrointestinal (GI) Candida commensals may be a major source of invasive candidiasis and a hidden reservoir of antifungal resistance. Candida glabrata resistance rates have increased greater than those of other species. Here, we present a C. glabrata GI colonization model to explore how antifungal drugs affect resistance acquisition and systemic breakthrough infections. METHODS: Immunocompetent mice were treated with antibiotics to clear native GI bacteria and then inoculated via oral gavage with C. glabrata. Fecal samples were collected throughout the study to assess fungal GI colonization. Daily administration of caspofungin (CSF; 5 or 20 mg/kg i.p.), chitin synthase inhibitor nikkomycin Z (Nz; 100 mg/kg oral), or saline was initiated on day 3 post inoculation. CSF-resistant colony frequencies were determined through selection of fecal samples on CSF-supplemented media, and FKS mutations were identified using the newly developed molecular beacon diagnostic assays. Dexamethasone was administered to induce immunosuppression. Upon completion of the experiment, blood, and organs were harvested and yeast burden levels determined. RESULTS: Daily therapeutic dosing (5 mg/kg) of CSF resulted in no reduction in fecal burdens, little resistance (0–10%), and organ breakthrough rates similar to control groups. Treatment with high dose (20 mg/kg) CSF caused a 2.5-log decrease in average burden, yet high levels (10/10 mice) of resistance (fks1/2 mutants) were observed following 5–9 days of treatment. Although breakthrough rates decreased in this group, yeast recovered from organs contained fks mutations. The largest reduction (3 log) in GI burdens was obtained within 3–5 days of high dose CSF plus Nz (100 mg/kg; oral) treatment. However, echinocandin resistance was again observed from all mice (10/10) following 5–7 days of treatment. Treatment with the therapeutic dose plus Nz left GI burdens unchanged, but did significantly reduce organ breakthrough rates (20%; P < 0.05). CONCLUSION: We have developed a C. glabrata GI colonization and dissemination model. Systemic breakthrough depends on both gut C. glabrata population composition and serum/tissue drug level. DISCLOSURES: D. Perlin, Merck: Grant Investigator and Scientific Advisor, Research support Pfizer: Grant Investigator, Research support. Astellas: Consultant and Grant Investigator, Research support and Speaker honorarium. F2G: Consultant, Consulting fee