(700e) Model Based Experimental Design and Parameter Estimation for An Industrial Hydrophobic Interaction Chromatography Step From a Quality by Design (QbD) Perspective

Sørensen, E., University College London
Bracewell, D., University College London
Salm, J., Pfizer, Inc
Lyons, J., Pfizer Inc

Integrated modelling and experimentation is an exciting approach to the design of chromatographic processes for the manufacture of therapeutic proteins. The approach has significant advantages over traditional process development methods, especially considering the desire to move to a Quality by Design (QbD) processing paradigm. The development of a mechanistic model derived from first principles is, however, often experimentally expensive as well as time consuming which is a major reason for their limited use in a biopharmaceutical industry dominated by time to market pressure.

In this work we begin to address these issues when considering a hydrophobic interaction step from an industrial purification process during the manufacture of a commercial therapeutic protein. Our work considers model based experimental design to optimally direct high throughput experimentation and column studies for the parameter estimation of a detailed first principles chromatography model.

The proposed methodology determines the experimental conditions, parameters to measure, and when to measure those parameters, in order to maximise the information gathered from the measurements in the context of utilisation for estimating model parameters. The methodology enables a model of the desired level of accuracy to be developed quicker than otherwise possible, and at minimum experimental cost. The results can be used to complete a range of highly desirable design tasks, such as resin screening, design space exploration and process optimisation, all key to the successful implementation of a Quality by Design (QbD) process development strategy.