B107-21 Translational Assessment of Inhalable Therapies Using Multi-Model Integration

Abstract Rationale Inhalation-based delivery of anti-infectives—including RNA therapeutics, antivirals, and antibiotics—offers a targeted strategy for treating respiratory infections directly at the site of disease. To accelerate development and reduce reliance on animal testing, predictive models for early feasibility testing are essential. This study presents a translational platform that integrates in vitro, ex vivo, and in vivo approaches to assess efficacy, toxicity, biodistribution, and pharmacokinetics of novel inhalable anti-infective candidates efficiently. Methods The platform integrates complementary models, each selected for its specific predictive strength. For human proof of concept and to define a therapeutic window, a Precision Cut Lung Slice (PCLS) model was used as an ex vivo organotypic infection model for investigation of the host-pathogen interaction in situ. The Isolated Perfused Rat Lung (IPL) provides a fully intact organ system for analyzing organ-level toxicity and systemic absorption of inhaled compounds, supporting early pharmacokinetic profiling and dose translation. Local cytotoxicity was assessed using PRIT ExpoCube® air-liquid interface cultures of human bronchial epithelial cells, offering high-throughput and mechanistic insights. For selected compounds, corresponding in vivo data—such as from a 28-day inhalation study with Nafamostat—demonstrated strong agreement with ex vivo predictions. All data streams were integrated through dosimetric and pharmacokinetic modeling approaches to support cross-model dose translation and enhance human relevance. Results Across compound classes, the platform generated predictive data that aligned with in vivo outcomes. For antibiotics such as Ciprofloxacin and Tobramycin, effective concentrations in PCLS matched known MIC values, and IPL-based permeability (Papp = 4*10-7 cm/s) and distribution (1:4.6 for ciprofloxacin) measurements enabled prediction of systemic exposure consistent with published human pharmacokinetics. RNA-based therapies showed dose-dependent delivery to bronchial (8%) and pulmonary cells (4%) in vivo in agreement with ex vivo IPL findings. Lipid nanoparticles were confirmed as key contributors to respiratory toxicity, with observed thresholds closely matching those derived from ex vivo models. For Nafamostat, a 28-day inhalation study in rats demonstrated adverse effects at 0.5 mg/kg, validating LOAEL estimates from IPL and epithelial cell models. These results confirm the translational value of the integrated approach for early feasibility testing of inhalable anti-infectives. Conclusion The integrated platform enables early, quantitative and predictive assessment of inhalable anti-infectives across compound classes by combining human-relevant ex vivo models with complementary in vitro and in vivo approaches. It reliably anticipates efficacy, toxicity, and pharmacokinetic behavior, supporting dose selection and therapeutic window definition in the preclinical phase, thereby contributing to faster and more efficient drug development. This abstract is funded by: German Federal Ministry of Education and Research