Advances in Crystallization and Precipitation Technologies for Biotherapeutics

Modeling approaches can help develop and optimize precipitation and crystallization for a wide range of therapeutical products, such as monoclonal antibodies and Adeno-associated viruses.

Separation is increasingly becoming a bottleneck during the process development and production of biotherapeutics. Alternative separation technologies, such as crystallization and precipitation, have the potential to significantly reduce process development time and costs. The characteristics of biotherapeutics pose unique challenges for the development of separation technologies. Mechanistic modeling and experimental results demonstrate the promise of novel downstream technologies, which are supported by model-based optimal design and control. Recent research has shown the viability of these technologies for multiple biotherapeutics, including monoclonal antibodies (mAbs), a protein sub-unit vaccine, and AAV-gene therapies.

Conventional preparative chromatography has been the workhorse in downstream purification processes for therapeutic proteins over the decades. This method is favored for its high separation efficiencies, facilitating the recovery of biomolecules with both high yield and purity. However, preparative chromatography comes with its own set of challenges, primarily due to an inherent mass transfer limitation (1). This limitation becomes more pronounced, particularly when dealing with high product concentrations, which can act as a bottleneck in downstream processing. Additionally, the high cost of chromatographic adsorbents, coupled with linearly scaling operating costs, further accentuates the need to explore alternative purification technologies (1).

One promising alternative is preparative crystallization/precipitation, which presents significant economic advantages over chromatography. This method eliminates mass transfer limitations and requires low-cost equipment and consumables, proving effective and cost-efficient for achieving high purity and large-scale production (2). Notably, preparative crystallization offers distinct advantages over conventional formulations, such as the ability to prepare highly concentrated doses with lower viscosity and a potentially better patient experience with consistent controlled release of activity; additionally, this separation method helps create amorphous lyophilizates and aqueous solutions that feature enhanced stability and higher purity of the crystalline biomolecule (2).

The adoption of molecular and process modeling may pave the way for the industrial application of crystallization/precipitation as a purification process for biopharmaceuticals. Within the framework of quality by design (QbD) initiatives, modeling is increasingly integral in biopharmaceutical process technology as a way to ensure efficacy before costly and lengthy experimentation. Mathematical models help reveal the effects of critical material attributes (CMA) and critical process parameters (CPP) on product quality and productivity (3). The benefits of modeling include lower experimental effort, improved process understanding, and a rational foundation for decision-making. Additionally, model-based methods can play a vital role in process monitoring, optimization, and control when used in conjunction with process analytical technology (PAT). Given that typical protein crystallization/precipitation experiments are based on a micro-batch scale, modeling plays an important role in the process scale-up.

This article delves into efforts to develop platform technologies for protein crystallization/precipitation, emphasizing the critical role of modeling in improving process understanding. The discussion begins by presenting the fundamental theory of crystallization/precipitation and the challenges associated with it. Subsequently, the article reviews case studies focusing on the crystallization of monoclonal antibodies (4), showcasing successful applications of molecular modeling to assess the ease of crystallization, and employing a systematic approach using a droplet-based automated evaporative system for optimal scale-up. Finally, the article demonstrates an integrated experimental and modeling approach for purifying adeno-associated virus (AAV) vectors, highlighting the unique challenges and potential benefits of this method in the context of AAV vectors.