(160h) Surrogate-Based Feasibility Analysis for the Identification of Design Space of Multicolumn Counter-Current Continuous Protein a Chromatography

Monoclonal antibodies (mAbs) are among the most critical biopharmaceutical products¹, and the most widely used method for biological production of mAbs is operated in batch operation mode^{2, 3}. However, due to the fast expansion of market demand, continuous production becomes a promising alternative to produce mAbs, which has various advantages including higher and constant product quality, increased productivity and yield, smaller footprint, and rapid capacity adjustment⁴. The commercial mAb production process can be divided into upstream process for protein production and downstream processes for protein purification⁵. With rapid improvements in continuous upstream processes like the development of perfusion bioreactors, continuous downstream processes have become the technical bottleneck in achieving the integrated continuous bioprocess^{6, 7}. In particular, primary capture, which is performed to remove most of the process-related impurities like culture media components and host cell proteins, has attracted enormous attention since this step accounts for nearly 33% of downstream process cost⁸. Protein A affinity chromatography is a benchmark tool for primary capture because of its high selectivity and efficiency⁹, but its development is mainly impeded by the high resin costs. To address this issue related with the high costs, continuous chromatography like multi-column periodic counter-current chromatography (PCC) has been developed, which can efficiently improve productivity, increase resin capacity utilization, and lower buffer consumption^{10, 11}.

In order to gain an insightful understanding about process variable effects on product quality, Quality by Design (QbD) is initiated by the FDA to increase patient safety by incorporating product quality into process design¹². Design space can help determine the operational ranges of a process to ensure product quality and help understand the main effects and interaction effects of process variables on product quality. Besides using design of experiments (DOE) to identify design space¹³, modeling can be also employed to reduce the experimental cost and explore the impact of critical process parameters (CPPs) on the critical quality attributes (CQAs). For instance, the design space for continuous frontal chromatography process¹⁴ and twin-column MCSGP¹⁵ were identified based on mechanistic model by sampling the full operating space, but this approach would result in a tradeoff between model accuracy and computational cost. Increasing the sample size may result in improved accuracy of the identified design space, but it would increase the sampling cost simultaneously.

In this work, surrogate-based feasibility analysis with adaptive sampling approach^{16, 17} is developed to replace the mechanistic model to identify the design space of multi-column PCC in an effort to balance model accuracy and computational complexity. The general rate model with Langmuir isotherm to describe the adsorption process is used as mechanistic model. Mass transfer and adsorption parameters are first estimated by fitting batch breakthrough curves at different flowrates and different feeding concentrations. The predictive model for continuous twin-column PCC is thus generated and validated by comparing the simulated column performance (e.g., productivity, yield, resin capacity utilization, buffer consumption, and outlet concentration) with the experimental results. Surrogate-based feasibility analysis with adaptive sampling method is employed to identify the design space of this system in order to maintain high accuracy but reduce the computational cost. More specifically the surrogate model is established based on the relationship between inputs (process variables including interconnected loading time, interconnected flowrate and batch flowrate) and the outputs (the maximum value among productivity, yield and capacity utilization constraints) obtained from the developed mechanistic model. Using this approach, the change of design space with respect to different process variables is investigated based on active set strategy to determine and thoroughly characterize the changes in column performance. In addition, the impacts of recovery-regeneration time and product quality constraints on the design space are also studied. Based on the established predictive mechanistic model, the impacts of design variables (i.e., three-column PCC system) on the design space is also analyzed and the difference of design space between twin-column and three-column PCC is presented.

References

1. Lu, R. M.; Hwang, Y. C.; Liu, I. J.; Lee, C. C.; Tsai, H. Z.; Li, H. J.; Wu, H. C., Development of therapeutic antibodies for the treatment of diseases. J. Biomed. Sci. 2020, 27 (1), 1.
2. Zydney, A. L., Perspectives on integrated continuous bioprocessing—opportunities and challenges. Curr. Opin. Chem. Eng. 2015, 10, 8-13.
3. Croughan, M. S.; Konstantinov, K. B.; Cooney, C., The future of industrial bioprocessing: batch or continuous? Biotechnol. Bioeng. 2015, 112 (4), 648-51.
4. Yang, O.; Qadan, M.; Ierapetritou, M., Economic Analysis of Batch and Continuous Biopharmaceutical Antibody Production: A Review. J. Pharm. Innov. 2019, 14, 1-19.
5. Tripathi, N. K.; Shrivastava, A., Recent Developments in Bioprocessing of Recombinant Proteins: Expression Hosts and Process Development. Front. Bioeng. Biotechnol. 2019, 7, 420.
6. Shi, C.; Gao, Z. Y.; Zhang, Q. L.; Yao, S. J.; Slater, N. K. H.; Lin, D. Q., Model-based process development of continuous chromatography for antibody capture: A case study with twin-column system. J. Chromatogr. A 2020, 1619, 460936.
7. Somasundaram, B.; Pleitt, K.; Shave, E.; Baker, K.; Lua, L. H. L., Progression of continuous downstream processing of monoclonal antibodies: Current trends and challenges. Biotechnol. Bioeng. 2018, 115 (12), 2893-2907.
8. Yang, O.; Prabhu, S.; Ierapetritou, M., Comparison between Batch and Continuous Monoclonal Antibody Production and Economic Analysis. Ind. Eng. Chem. Res. 2019, 58 (15), 5851-5863.
9. Ramos‐de‐la‐Peña, A. M.; González‐Valdez, J.; Aguilar, O., Protein A chromatography: Challenges and progress in the purification of monoclonal antibodies. J. Sep. Sci. 2019, 42 (9), 1816-1827.
10. Sun, Y. N.; Shi, C.; Zhang, Q. L.; Yao, S. J.; Slater, N. K. H.; Lin, D. Q., Model-based process development and evaluation of twin-column continuous capture processes with Protein A affinity resin. J. Chromatogr. A 2020, 1625, 461300.
11. Angarita, M.; Müller-Späth, T.; Baur, D.; Lievrouw, R.; Lissens, G.; Morbidelli, M., Twin-column CaptureSMB: A novel cyclic process for protein A affinity chromatography. J. Chromatogr. A 2015, 1389, 85-95.
12. St Amand, M. M.; Hayes, J.; Radhakrishnan, D.; Fernandez, J.; Meyer, B.; Robinson, A. S.; Ogunnaike, B. A., Identifying a robust design space for glycosylation during monoclonal antibody production. Biotechnol Prog 2016, 32 (5), 1149-1162.
13. Xiang, H., Establishing Process Design Space for a Chromatography Purification Step: Application of Quality-by-Design Principles. BioPharm. International 2015, 28 (11).
14. Vogg, S.; Muller-Spath, T.; Morbidelli, M., Design space and robustness analysis of batch and counter-current frontal chromatography processes for the removal of antibody aggregates. J. Chromatogr. A 2020, 460943.
15. Steinebach, F.; Krättli, M.; Storti, G.; Morbidelli, M., Equilibrium Theory Based Design Space for the Multicolumn Countercurrent Solvent Gradient Purification Process. Ind. Eng. Chem. Res. 2017, 56 (45), 13482-13489.
16. Wang, Z.; Escotet-Espinoza, M. S.; Ierapetritou, M., Process analysis and optimization of continuous pharmaceutical manufacturing using flowsheet models. Comput. Chem. Eng. 2017, 107, 77-91.
17. Rogers, A.; Ierapetritou, M., Feasibility and flexibility analysis of black-box processes part 2: Surrogate-based flexibility analysis. Chem. Eng. Sci. 2015, 137, 1005-1013.