More than 85% of novel biopharmaceuticals that have received approval in recent years have been produced via mammalian cell culture. Chinese hamster ovary (CHO) cells remain the top large molecule expression system, and 89% of large molecule active pharmaceutical ingredients (APIs) are produced in this manner.
Biologics contract development and manufacturing organizations (CDMOs) will continue to ramp up capacity for the production of mammalian and microbial programs while also investing in key advancements that continue to transform the industry. Three fascinating areas of innovation include the merging of machine learning (ML) and Design of Experiments (DoE) to improve product quality and process development efficiency, bioseperation system advancements, and developments helping to alleviate the labor shortage in the bioprocessing industry.
Science, including biologics development and manufacturing, is an iterative process. DoE is a powerful data collection and analysis approach that began gathering momentum in the 1930s. Although DoE is not new, the convergence of robust data processing technologies capabilities, accessibility, and affordability within the past fifteen to twenty years has dramatically increased the adaptation of this methodology.
Expressly, sophisticated DoE software packages have provided practical, rapid, and low-cost ways to explore the multifactor dynamics that affect every aspect of biologics process development, including upstream and downstream processes.
Digital technologies are better at creating the many experimental combinations necessary for a robust DoE approach, and DoE technology facilitates the removal of human emotion, bias, and faulty interpretation. The benefits of this experimentation efficiency cannot be overstated; however, the next wave of advancements will be facilitated by applying artificial intelligence (AI) and ML technologies to process development and continuous manufacturing improvement efforts.
Machine learning and Design of Experiments teaming up to improve biologics process development
Work is underway to converge ML and DoE in two ways. The first application uses DoE as a systematic approach for selecting the best combination of parameters for ML models. The DoE methodology speeds up the interpretation of each parameter’s effects on the algorithms’ performance as well as the ML training process.
The second, and arguably most exciting application, is using ML to analyze and act upon DoE data. Specifically, biologics CDMOs are exploring ways to use ML to model complex relationships housed within DoE data efficiently.
Many in the industry envision ML-based systems replacing DoE approaches; however, significant limitations must be addressed. Specifically, today’s ML requires vast sums of continuously generated data. Many biologics CDMOs are generating enough data within their production processes to consider applying ML to manufacturing process improvement initiatives. However, the quantity of data needed for ML-based systems to replace DoE for process development will likely not be available any time soon. Therefore, rather than replacing DoE with ML, researchers should use DoE and ML together to improve and speed up biologics process development.
Until fairly recently, biomanufacturers faced limited separation process options. Choices were primarily confined to technologies such as fixed-bowl and disk-stack stainless-steel centrifuges and methods such as depth filtration. Although these systems worked, biologics contract development and manufacturing organizations were impeded by the expensive and time-consuming cleaning validation procedures for changing production lines.
The development of GMP-compliant single-use continuous centrifuges and advanced filtration materials have provided needed breakthroughs. Like all unit operations using single-use systems, single-use centrifuges enable the quick production line changeover that multiproduct biologics CDMOs and their clients need.
Filtration advancements increase efficiency for biologics CDMOs
Until recently, extensive filter trains have required considerable amounts of water to flush the filters, preparing them for operation. Additionally, these large filter trains—often three or four stacks with seven to ten filters each—substantially increased the overall processing time required for CDMOs in the US and around the world. However, advancements in filtration technology and materials have reduced the necessary filtration surface area, lessening required water usage and shortening filtration processing timelines.
Depth filtration as a vital biopharmaceutical separation technology is not going away anytime soon. However, continuous single-use centrifuges coupled with modern filtration materials increase the speed of recovery processes, reduce required energy and water, speed up processing times, and often improve product recovery.
Biopharmaceutical CDMOs face talent shortages linked to broader labor market trends, as in many other markets. Notably, there has been a 20% increase in demand for STEM-related roles across the life sciences industry in the United States, and the pool of talent is considerably smaller than this demand, making it challenging to recruit scientists, operators, and other technical talent.
Given this trend line, the advancement of digital tools and solutions has become more critical. By adopting digital and analytical tools at scale, innovative biologics CDMOs are reducing the effects of labor shortages through highly efficient planning, advanced automation-enabled biopharmaceutical manufacturing, proactive risk management, and remote performance monitoring. Additionally, as discussed above, they are leveraging ML and AI to improve biologics process development efficiency.
Biologics CDMOs are also thinking creatively about recruiting, training, and retaining top talent. The current pool of experienced biotech candidates is not large enough to meet the industry's growing needs. That's why CDMOs and innovators are considering new, untapped talent markets, such as veterans looking to transition back to civilian careers. The highly regulated, discipline-driven biotechnology industry is a natural fit for military veterans. Additionally, there is an increasing need to partner with local and regional educational institutions to make students aware of the opportunities in the biologics industry. Internship programs for high school and college students and curriculum development partnerships with universities are just two ways CDMOs can help grow the pipeline of talent needed to manufacture the next generation of biologics.
The Scorpius senior management team and technical scientists and engineers in the organization have many decades of industry experience in all aspects of the market. Scorpius’ team members are energized by ongoing efforts to advance how a modern biologics CDMO can serve the current and future needs of the market.
As advancement continues, Scorpius BioManufacturing’s primary mission is to accelerate innovators’ products through their development journey to market. By embracing and developing progress, Scorpius is creating more cost-effective paths to commercialization.