In the race for investigational new drug (IND) submissions, innovative molecular entities, such as bispecific antibodies, antibody fragments, and fusion proteins, demand rapid yet robust process development. Process optimization, regulatory alignment, and risk mitigation are critical for drug developers wanting to compress project timelines as much as possible. To rapidly streamline the path from lab to market, engaging with a contract development and manufacturing organization (CDMO) with a comprehensive understanding of in silico and physicochemical analyses and a wealth of process development experience is vital for development success.
Continuing Trend of Expedited Biologic Drug Development
Stakeholders, including drug development companies, regulators, and governments around the world, have recently focused on finding ways to drive down the cost and time required to develop new drugs to increase accessibility to therapeutics, including both generic and novel products. At the same time, governments and payors expect new drugs to provide clear advantages over existing medicines in terms of increased efficacy, reduced costs, or greater ease of use and convenience for patients.
Being first to market with a new drug is essential to claim and maintain market share. Consequently, many developers have focused on specialized programs that qualify for expedited approval pathways based on experience, which could reduce development timelines while lowering the risk of failure. As a result, developers seek CDMOs with accumulated experience and proven track records of success.
Expediting Development Creates Challenges
Comprehensive characterization of complex processes and products, including all related impurities, is challenging under expedited timelines. Structural similarities between a desired drug product and multiple product-related impurities complicate efforts to identify and distinguish between them. Moreover, the industry is progressing toward more complex molecular structures beyond those of the traditional monoclonal antibody (mAb). This indicates that the process development path is becoming more intricate, demanding greater effort for success. For such complex molecules, multiple orthogonal analyses using state-of-the-art techniques must be leveraged to ensure product quality, safety, and efficacy. In many cases, fit-for-purpose methods are not yet available, as they require a deep understanding of development within the constraints of accelerated timelines.
In essence, developers working with molecules with such complex structures need diverse technical capabilities and specialized expertise, along with proficient project management experience.
Focusing on Hurdles in Downstream Process Development
Process development for molecules with varying characteristics requires a customized approach. Even mAbs, for which platform processes are widely implemented, require non-trivial process optimization efforts for each molecule.
Downstream process (DSP) aims to achieve both high yields and purity. Platform mAb processes serve as an initial process development framework for many antibody derivatives, including multi-specific antibodies, antibody fragments, fused antibodies, and ADCs, however, the unique characteristics of these complex molecules frequently require substantial modifications from this foundation. Entirely novel DSPs may be necessary to accommodate the specific needs of each antibody variant and to ensure the production of high-quality biologics. Whether building on platform processes or engineering novel processes, the approach must be rigorous and include a series of iterative design of experiments (DoE) for resin and parameter screening and optimization.
Despite these complexities, CDMOs are pressed to achieve rapid IND approvals by condensing the process development timeframe. Moreover, they strive to minimize costs by scaling down the production of GMP-compliant materials, prioritizing high-yield process development and insisting on advanced purification processes to remove early-stage impurities, such as smaller molecular fragments that may impact the drug’s potential. This presents an ongoing challenge: to balance the need for swift process development to adhere to tight IND submission schedules with the imperative to maintain high yield and uncompromised quality.
Streamlining DSP Development Through Operational Excellence and Molecular Insight
Samsung Biologics adopts a DSP strategy that marries molecular insight developed from in silico and physicochemical analyses with a track record of developing over 100 molecules and operational efficiency aimed at expediting the drug development timeline—particularly for complex molecules without established platform processes. The DSP flow can be set at an early stage based on historical data and understanding the molecular characteristics gained from analyses (Figure 1). The seamless orchestration of teams of experts and robust, high-throughput platforms for processes and methods and the effective management of critical raw materials allow for early predictions of the manufacturability of a biotherapeutic and preparations for tech transfer and ensure that quality and supply remain consistent and uninterrupted.
Figure 1. Downstream process workflow.
The company’s robust DSP strategy is founded on a systematic approach to understanding and evaluating molecular characteristics in combination with historical data to predict the DSP flow and optimize resin screening.
An in silico approach to predicting column performance for protein purification based on protein structure is crucial for efficiencies. A comprehensive understanding of the target protein’s folding, stability, and aggregation propensities provides insight into its interactions with chromatography resins under specific process parameters, such as pH and ionic strength. By calculating characteristics like net charge differences between the target protein and impurities based on historical DSP data, experts can precisely select an appropriate chromatography approach (e.g., ion-exchange, hydrophobic interaction, or mixed-mode chromatography) and a resin for initial DoE testing (Figure 2).
Figure 2. In silico prediction using structure and chemical bonds. If the protein structure indicates a difference in chemical binding force, column performance can be predicted.
The company can complement the in silico predictions with physical and chemical property tests to reinforce resin selection efficiency. DSP scientists can leverage high-throughput tests that evaluate solubility by monitoring levels of aggregation at different levels of hydrophobicity to pinpoint the sensitivity of the molecule and implement the selection of a hydrophobic resin (Figure 3). A molecule with relatively greater sensitivity to hydrophobicity would prompt consideration for a weak hydrophobic condition for chromatography. By leveraging accumulated project data for both the in silico predictions and chemical testing results, scientists can achieve a refined selection of resins that enhance recovery of the target protein, raising the probability of selecting the ideal resin for maximal yield and reducing the timeline (Figure 4).
Figure 3. Physiochemical property test. Hydrophobic resin can be selected based on test results.
Figure 4. Resin screening for the investigational new drug process. Historical in silico and chemical testing increases the chance of selecting the optimal resin for high yield and timeline reduction.
The company demonstrated the effectiveness of applying its strategies by comparing scenarios with and without the resin screening strategy. In the case of Fc fusion during resin selection, the results reveal a significant improvement in yield and reductions in low-molecular-weight species as a result of implementing the strategy (Figure 5).
Figure 5. Resin screening to optimize yield.
In addition to selecting the appropriate resin, determining the optimal conditions for performance is also a crucial consideration. In another Fc fusion project, following the identification of the optimal resin, the group compared two conditions at varying pH levels. The group found that a higher pH increased step yield and reduced high-molecular-weight species. Varying the conditions while using the same resin resulted in notable differences in performance (Figure 6).
Figure 6. Optimization through pH control.
The final example, involving a bispecific antibody developed using Samsung Biologics’ proprietary platform S-DUAL™, maximized purity levels to 99% (Figure 7). Hence, not only was purity maximization possible, as demonstrated by the successful resin selection, but the company was also able to thoroughly evaluate and optimize purity for different projects by incorporating DoE into its practices (Figure 8).
Figure 7. Optimization for purity.
Figure 8. DSP optimization. Purity and yield % were enhanced through downstream process optimization. Resins and buffer compositions during washes, elution, and flowthrough were evaluated and optimized via design of experiments.
Supporting Expedited DSP Development at Samsung Biologics
Emphasizing these operational aspects and utilizing an approach grounded in molecular understanding helps set a new standard in DSP development to expedite the timeline without compromising quality and efficacy. Samsung Biologics leverages its accumulated experience and knowledge to support the need for rapid and reliable process development methodologies tailored to meet the specific demands of individual biologics drug developers. As we endeavor to meet these global demands, the challenge of fully characterizing complex biological processes within tighter timelines necessitates innovative solutions to maintain the appropriate balance between speed and safety.
Although the compressed timelines for developing and approving two COVID-19 mRNA vaccines advocated by government and cross-industry collaborations have been the primary drivers for raising expectations for rapid development, many hope for the long-term re-evaluation of drug development and consistently reduced time to clinic and market. While such acceleration is not yet possible across drug development, evolving high-throughput methods and the growing use of artificial intelligence and machine learning throughout the drug development cycle show promise in achieving those goals.
