Bioprocessing of High Concentration Protein Solutions: Quality by Digital Design Approach

There is a need for underpinning research to support industrial development of novel protein therapeutics for more convenient delivery by subcutaneous injection (SC). This is an increasing priority for biopharmaceutical companies such that patients can administer the medicines at home, rather than having to visit hospital for a lengthy infusion. The challenge for bioprocessing research is to dissolve the dose of protein required in a small volume, usually 1 ml, that can be self injected. The protein therefore must be soluble up to 300 mg/ml, and it is desirable that the liquid can be stored at 2-8 C for 2 years or more without precipitation, aggregation or other instability. In addition, the liquid must not be too viscous, otherwise the injection will require too high a pressure or may take too long to administer. There is also a need to prevent damage to the protein during the process of forcing the liquid through a narrow needle, into the tissue under the skin.

The proposed research will develop methods for use by industry to screen protein formulations for viscosity and other flow properties, using small quantities of protein. This will enable methods for viscosity reduction to be developed. It is known that similar proteins differ widely (by a factor of two or more) in their viscosity at similar concentrations, and that alterations in co-solvent can reduce the viscosity of a formulation. To achieve this, we propose to apply comprehensive rheological characterisation, RheoChip rheometry, and advanced modelling as a platform, which can be used by industry to select the protein and formulation for development of the final dosage form, at an earlier stage than it is possible today. This should save time and cost in development of many new protein medicines. The research will build on existing methods, which are already well established for rheological characterisation of water soluble polymers and BSA solutions, and adapt and apply them to the bioprocessing and injectability of high concentration protein biopharmaceutical solutions. Comprehensive rheological characterisation of protein solutions has not yet been published. In addition, there is the potential for this new knowledge to be applied in industry to improve the production of biopharmaceutical proteins, as high concentrations may be reached during bioprocessing, e.g. freeze drying, tangential flow filtration (TFF) etc. and there can be difficulties in processing viscous solutions, e.g. nanofiltration for virus removal may be impractical. The deliverables of the project will be the form of instrumentation, rheological characterisation methods more relevant than current viscosity measurement, and computational tools.

The project has five work packages (WPs). WP1 and WP2 will focus on development of new enabling technologies. The output of WP1 will be the first high throughput characterisation platform for screening protein formulations under the flow conditions encountered in bioprocessing while requiring minimal sample. WP2 will construct a computational platform for predictive modelling of concentrated protein fluid flows. WP3 and WP4 will critically validate these enabling technologies using both model and industrially relevant protein solutions under the complex flows, including TFF and SC injection. The output will be an integrated approach for design and optimisation of (nonlinear) scale-up protein production, based on high throughput rheological data obtained from Rheo-chip and predictive modelling of protein flows and protein stability during processing. WP5 will correlate the rheological properties and flow behaviour of concentrated protein solutions with the effects of excipients and/or formulation conditions on the conformational stability and self-association in dilute solution. This will help to establish the molecular determinants of high viscosities and flow induced protein aggregation leading to rational design of high throughput screens.

Future trends in treating various chronic diseases with recombinantly-produced therapeutic proteins (such as monoclonal antibodies) require frequent and high dose of active protein ingredient in a small volume of liquid (e.g. >100mg/ml for subcutaneous (SC) injections using prefilled syringe or auto-injection device). SC injections will improve the convenience of administration of products for patients, and reduce healthcare costs by greatly reducing the time for the patient to stay in hospital. The convenience of SC administration drives for the development of concentrated protein medicines and poses fundamental challenges for bioprocessing in terms of solubility, stability, manufacturability and drug deliverability.

We will address those fundamental challenges by translating a novel Rheo-chip technology and state-of-the-art numerical techniques of computational rheology into biopharmaceutical industry. Rheo-chip will enable high throughput rheological characterisation of protein formulation, fast evaluation of their stability under industrial and clinical relevant flow conditions using small sample volumes. It will become a unique formulation-screening platform for evaluation of the stability, manufacturability and drug deliverability of protein solutions. The simulation platform can account for spatial distribution of protein aggregates in complex flows. Its uniqueness is to link molecular scale protein-protein interactions, meso-scale of protein aggregations with the macro-scale nonlinear flow behaviour of protein solutions. By integrated with the Rheo-chip platform, it will enable optimal design of unit operations and manufacturing scale-up as well as syringe design for optimal delivery of concentrated protein products from very early stages of drug development that inherently embrace Quality by Digital Design.

The proposed project is in response to a Call for Proposal by the Bioprocessing Research Industry Club (BRIC). It is based on extensive consultation with BRIC industrial partners on their unmet needs, and will address the fundamental challenges of manufacturing concentrated protein medicines by translating a novel Rheo-chip technology and state-of-the-art numerical techniques of computational rheology into biopharmaceutical industry. The project has relevance to 4 out of the 5 of the BRIC2 Research Priority Areas, including
1) Bioprocessing research challenges for protein products;
2) High-throughput bioprocess development;
3) Effective modelling of bioprocess;
4) Robust and effective analytics for bioprocessing.
It also addresses two key BRIC2 Business Drivers "the direct commercial and competitive value to existing companies from decreasing the time, cost and risk of product development" and "the reduction in capital investment magnitude and risk to existing companies coming from the ability to design intensive, modular and predictable processes that inherently embrace 'quality by design'."

In particular the high throughput Rheo-chip based characterisation platform will be available to fulfil the unmet needs in bioprocessing industry for rapidly screening the manufacturability of a large number of protein formulations in an early stage of drug development. The small sample requirements of Rheo-chips and the measurement protocols developed from the project will permit rapid screening of the design space, including variables such as protein concentration, temperature, solvent conditions (buffer type, excipient, ionic, strength, pH) and in the presence of air interfaces. In-situ rheological and birefringence/velocity field characterisation of protein solution in Rheo-chips will reveal the dynamic mechanisms of the self-assembly and disassembly of the proteins, as well as establish the hydrodynamic conditions which can trigger the aggregation or instability of proteins, hence the link between protein aggregation and stability on storage with the deformation history and interfacial exposure of the sample can be identified. By critical correlation analysis between Rheo-chip results and measurements of protein-protein interactions, we will search for strongly correlated signature between the nonlinear rheological and flow properties of concentrated protein solutions and protein-protein interactions in dilute solution, hence to explore an approach of rationally designing screens to minimize sample use, which combine a small number of rheological measurements on concentrated protein samples with a larger number of measurements of protein-protein interactions in dilute solution.

The predictive simulation platform will be available for design and optimisation of scale-up manufacturing and subcutaneous injection, hence to maximise protein yield and stability, and to accelerate development with reduced costs and risks. The uniqueness of the proposed micro- & macro-coupling computational model is to link molecular scale protein-protein interactions, meso-scale of protein aggregations with the macro-scale nonlinear flow behaviour of protein solutions. Large scale modelling of concentrated protein solutions in ultra-TFF operations, simulations of the turbulence enhancements, sensitivity analysis will allow us to identify the properties of concentrated protein solutions that control fouling, flux and retention during filtration. Therefore the impact of the project on UK and international companies developing and manufacturing therapeutic proteins are expected to be very significant. The ultimate beneficiaries are patients, who will receive treatment in a less invasive and more convenient form. In addition there is the potential of cost savings for healthcare providers such as the NHS when patients are able to self inject subcutaneously, rather than visit hospital for intravenous injection.