Ultrafiltration of Highly Concentrated Monoclonal Antibody Solutions

Open Access
Author:
Binabaji, Elaheh
Graduate Program:
Chemical Engineering
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
January 15, 2015
Committee Members:
  • Andrew Zydney, Dissertation Advisor
  • Andrew Zydney, Committee Chair
  • Themis Matsoukas, Committee Member
  • Manish Kumar, Committee Member
  • William O Hancock, Committee Member
Keywords:
  • Ultrafiltration
  • Memebrane
  • Concentration
  • Formulation
  • Monoclonal antibodies
  • Biotechnology
  • Purification
Abstract:
Monoclonal antibodies are currently the fastest growing segment of the pharmaceutical industry; these products are used in the treatment of a wide range of diseases including cancers and allergies. Antibody products are administered at high doses since the volume that can be delivered by injection is limited. Although ultrafiltration systems are used for final formulation of essentially all high value recombinant products, it is often challenging if not impossible to achieve the very high final formulation concentrations required for monoclonal antibody products. The overall objective of this thesis was to study and fundamentally understand the ultrafiltration behavior of highly concentrated monoclonal antibody solutions. This included: (1) evaluating the filtrate flux and maximum achievable antibody concentration during ultrafiltration of very high concentration (>200 g/L) antibody solutions, and (2) developing appropriate theoretical models to describe the ultrafiltration behavior in terms of independently measured biophysical properties of the antibody solution. The osmotic pressure of a highly purified monoclonal antibody at concentrations up to 250 g/L was evaluated over a range of pH and ionic strength, and in the presence of specific excipients, using membrane osmometry. These data were used to calculate the second and third virial coefficients. The second virial coefficients were in good agreement with independent measurements from self-interaction chromatography using a newly developed approach to evaluate the column dead volume. The second virial coefficient was positive under all conditions, consistent with repulsive electrostatic interactions between the positively charged antibody molecules. In contrast, the third virial coefficients were negative reflecting the presence of short-range attractive interactions between oppositely charged domains on adjacent proteins. Viscosity data were obtained over a wide range of protein concentrations, solution pH, ionic strength, and in the presence of different excipients. The concentration parameter in the viscosity correlation appeared to be well-correlated with the values of the third virial coefficient, consistent with the importance of short range attractive interactions. The viscosity and osmotic pressure data were incorporated in a modified concentration polarization model that accounts for the effects of intermolecular protein-protein interactions in the highly-concentrated antibody solutions as well as the presence of a back-filtration phenomenon arising from the large pressure drop through the tangential flow filtration module. This model was in very good agreement with experimental data and was able to correctly predict the maximum achievable protein concentration during batch ultrafiltration experiments conducted over a wide range of conditions. These results provide important insights into the factors controlling the ultrafiltration behavior for highly concentrated antibody solutions as well as a framework for the development of improved ultrafiltration systems for use in bioprocessing applications.