Quantitative prediction of inflammation from acute to chronic condition associated with inhalation and delivered ahead of time by lung animal-free in-vitro-learned digital twin (InFiniteLungDT)

InFiniteLungDT is a quantitative prediction tool for inflammation transitioning from acute to chronic conditions associated with inhalation. It is an animal-free, in-vitro-learned digital twin designed to predict long-term, time-evolved, dose-dependent inflammation triggered by inhaled poorly-water-soluble substances. InFiniteLungDT works by monitoring subcellular dynamics in vitro using a lung epithelium cellular cluster. Live-cell microscopy captures time-lapse images, which are then quantified into 17 time-dependent observables. A patented algorithm translates these into interaction couplings, defining the material's Mode-of-Action (MoA). This MoA is then used for in silico time-propagation to predict dose-dependent endpoints (neutrophil influx) and LOAELs for up to two years, extending beyond the in vitro observation timescale. Two cell lines are used, the murine epithelial lung tissue cell line (LA-4) and the murine alveolar lung macrophages (MH-S).

The intended purpose of the proposed method is to offer a resource-efficient alternative to animal testing, for regulatory applications (e.g. OECD TGs 403, 412, 413, and 452), delivering LOAELs for acute, sub-acute, sub-chronic, and chronic exposure scenarios. Additionally, it might support material developers in sectors such as advanced materials, nano and micromaterials, construction materials, and more, by facilitating early-phase pre-screening. The method is high-throughput compatible and can be scaled up for use by Contract Research Organisations (CROs).