Approach for scaling down pCO2 from clinical process to Ambr® 250 scale | Poster Board #189

Bioreactor scale down models (SDM) are necessary for process development and process characterization. Ambr® 250 fed-batch bioreactors have demonstrated large-scale reproducibility and significant process development time savings when used as SDM for GMP manufacturing. A late-stage IgG1 mAb produced in a CHO cell line has sensitivity to pCO2 concentrations and pH levels during both the growth and production phase that result in changes to process performance and critical quality attributes. These critical process parameters (CPP) are affected by both scale-independent factors (media pH, buffering capacity, and pH control setpoints) and scale-dependent process parameters (agitation, stir speed, and sparger design). Scale-independent process parameters can directly align with the large scale bioreactor control strategy. Scale-dependent parameters are generally chosen through scaling parameters such as maintaining power-to-volume (P/V), gas flowrate per unit volume (VVM), or the oxygen gas transfer coefficient (kLa) which are dominated by agitation and sparged gas flowrates. The main factors chosen in the study for controlling pH and pCO2 are agitation, air sparge, and CO2 gas flow rates in both open pipe sparger and the headspace. During the growth phase, air sparge has the greatest effect on pCO2 levels. During the production phase, VVM was dominated by the oxygen sparge rate as this cell line has a high oxygen uptake rate (OUR). The high VVM in the production phase decreased the pCO2 level from the target. Increasing the agitation to improve the oxygen gas transfer decreased the oxygen gas flowrate demands and subsequently increased the pCO2. Additional CO2 sparge has a greater effect than CO2 headspace gassing at increasing the pCO2. This work provided process parameter levers for modifying pCO2/pH at the Ambr® 250 scale which can significantly help develop more representative scale down models for more complex modalities and improve development timelines.