Scalability of mechanistic models for ion exchange chromatography under high load conditions

The majority of downstream processes (DSP) for biopharmaceuticals are based on chromatographic separation techniques. Future DSP development has to meet the evolving expectations of regulators, such as a mechanistic process understanding proposed within the QbD framework. The most sophisticated approach for developing mechanistic process understanding is the implementation of fundamental models for chromatography.

Monoclonal antibody in blue setting


To implement these models under real industrial conditions, the models need to be able to predict the highly overloaded conditions in preparative purification tasks. A second challenge is the transferability of model parameters between different column scales, which is important for process scale-up and the incorporation of small-scale data.

We present an industrial case study of ion exchange chromatography under high-load conditions used for the intermediate purification of a monoclonal antibody. At such high protein load densities, an unusual elution peak shape could be observed. This phenomenon cannot be modeled with the established equations for ion exchange chromatography, namely the steric mass action (SMA) isotherm. However, an activity-based extension of these isotherms could be applied successfully.


A systematical comparison of model calibrations in ion exchange chromatography using activity-based isotherms based on data collected from different chromatographic scales is presented. The case study covers a range of experimental systems from commonly used lab and small-scale columns, down to robotic high-throughput formats. We applied the scale invariant model parameters derived from one column scale to the other chromatographic scales for the prediction of the elution behavior.

This case study supports the fundamental assumption of in-silico scale-up and scale-down of chromatography, that only the fluid dynamics outside the pore system change. Once inside the pores, the same mechanism applies to robotic and production columns. Even for the observed complex adsorption behavior, the models calibrated from three gradients on a small scale were able to accurately predict the larger scale. A further scale-up to pilot and production scale is expected to work successfully.

The poster was presented at The Fourth International Conference devoted to high-throughput process development (HTPD); 10/2017; Toledo, Spain.