Model maths (Day 231)

With the growth in biochemical engineering and biopharmaceuticals, the challenge of bio separation is also being more widely addressed. In some mixtures, there are the issues of multi-component separations.

Biopharmaceuticals include proteins and other large molecules which may require complex chromatographic separations. Purification of biopharmaceuticals can account for 50-80% of the total cost of production and is often considered the bottleneck in the process.

The use of modelling and simulation in the development phase offers significant time and cost savings. This also includes efficient optimisation of the capital and operational expenditure.

A process to yield the product at the right purity, in an acceptable timescale and cost is essential. The US Food and Drugs Administration (FDA) and European Medicines Agency (EMA) regulations require reporting of the basic separations and purification procedures.

These must be provided to the US FDA early in the development process prior to clinical trials. This is aligned with the Quality by Design (QbD) approach used in the industry.

At large scale, the solution lies with chemical engineering. Through a combination of modelling and experimental work, separations can be understood and developed for optimal efficiency.

The increasing use of modelling has played an important role in managing costs and time in development.

Consideration of the multicomponent systems includes solvent properties, starting materials, impurities, by-products – their unique properties and how they interact with each other. In high value, small volume products such as biopharmaceuticals, chromatography is still widely used.

Some recently published work in Chemical Engineering Research and Design highlights a successful collaboration between Pfizer Biopharmaceuticals in the US and University College London (UCL) in the UK – with UCL's chemical and biochemical engineering departments both involved.

Researchers developed a mathematical model for the hydrophobic interaction chromatography of a protein mixture. The six component proteins were all similar with minor variations in the amino acid sequence resulting in subtly different forms.

Subsequent experimental work showed that not only was the separation of the components successful, but also the model accurately predicted the distribution and elution order.

This work is an excellent example of where mathematical modelling and experimental work carried out by chemical engineers is effective. In this case, the complimentary methods aided the development of a process that provides sufficient basic information to meet regulatory requirements in a cost- and time-efficient manner.

In the race to develop new medicines, refinement of a mathematical model enables more effective development.