(588cl) Beyond Trial-and-Error: An in silico Model for Accelerated Peristaltic Fill Recipe Development

Efficient and robust fill/finish operations are critical for drug product manufacturing, directly impacting throughput and product quality. Traditional approaches for developing a clinical or commercial filling recipe require several iterations to adjust for fluid properties, fill volume, and other SKU considerations. Fill recipe development often relies on at-scale characterization, making the effort particularly costly from the perspective of manufacturing capacity, materials, and speed-to-market. However, the use of in silico modeling for fill recipe development can accelerate such efforts and bypass traditional scale-up costs.

In this work, we describe the development, implementation, and application of an in silico model for peristaltic filling, a common filling technology. The model contains several components including predictions for pump speed and filling performance. This modeling approach enables in silico recommendations for optimal tubing size (i.e., fill sets), pump parameters (e.g., pump speed, drip retraction), and nozzle motion. Several use cases of the model will be presented to elucidate common tradeoffs and recipe decisions which would otherwise require line-time and trial-and-error experimentation.