Rapid bioprocess scale-up provides a competitive edge by reducing time to market. To meet market demand, a CHO-based manufacturing process was scaled up from a 2500 L bioreactor to six parallel 2000 L bioreactors. A Monte Carlo simulation-based model was developed based on the data set generated at the 2500L scale to provide in-depth predictions about process performance, robustness, and risks of different operating scenarios. This model allowed for informed decisions around equipment sizing, unit operation set points, and appropriate process limits for the production bioreactor and various downstream chromatography unit operations. Process expertise and data mining efforts identified and quantified sources of process variability. The process was then described using a mass balance model and simulated with randomly chosen process variables (e.g. unit operation yield) that adhere to the probability distributions derived from 2500L data set. The resulting model is capable of predicting the process output probability distributions under various potential conditions, including scenarios that maximized process yield, process robustness, or minimized changes in operating conditions during scale up from the 2500L process. These models were highly leveraged in the technical discussion on how best to scale-up the process.

Using a Monte Carlo simulation model to predict process performance during scale-up

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.