Validation of an In vivo Pulmonary Aspergillosis Model for Triazole Susceptibility Breakpoint Development

Use of mouse infection models for antimicrobial PK/PD analysis can assist in dosing regimen design and susceptibility breakpoint development. A major hurdle for clinical translation of in vivo study output is defining the model endpoint linked to clinical success. Validation of the in vivo endpoint requires a clinical data set composed of success or failure linked to MIC, dosing regimen, and if possible human pharmacokinetic measures. The present studies utilized a clinical library of 8 Aspergillus fumigatus strains in a mouse pneumonia model to define the endpoint associated with humanized treatment regimens of the triazole, posaconazole. This includes wild-type strains associated with successful treatment and strains with resistance mutations leading to elevated MICs and associated with treatment failure. We found humanized posaconazole exposures resulted in a net stasis or net decrease in organism burden in the animal model compared to the start of therapy or for all wild-type strains. However, a net increase in organism burden despite treatment with the humanized regimen was noted for strains with higher MIC values and defined Cyp51 mutations. The posaconazole AUC/MIC associated with a stasis endpoint in the mouse model was then utilized with susceptibility surveillance distribution data and human posaconazole population pharmacokinetics to perform probability of target attainment simulation. The output from these simulations observed >90% PTA for A. fumigatus strains with MICs of <= 0.5 mg/L in support of a susceptible breakpoint threshold.