Data at work: How advanced analytics are transforming cell and gene therapy manufacturing

Cell and gene therapy (CGT) has grown from niche research to a core force in biopharmaceutical innovation. By engineering living cells and genetic material to repair or replace defective biological mechanisms, these therapies can provide results that traditional drugs cannot. Still, manufacturing remains fragmented and highly variable, with most processes still relying on manual intervention and legacy systems, slowing progress in scalability.

The growing adoption of advanced analytics and artificial intelligence (AI) is reshaping the landscape by enabling data-driven process control, predictive manufacturing, and greater transparency throughout the development lifecycle. Together, these tools provide the foundation for scalable, repeatable, and compliant CGT manufacturing.

The expanding role of data in treatment development

Each stage of the CGT process, from cell collection to product release, generates large amounts of data. In the past, organizations stored much of their data in separate systems or tracked it manually, making it difficult to analyze or share.

Today, advanced analytics make it possible to combine and interpret this information in real time. By bringing together data from sensors, instruments and electronic records, the team can determine which factors have the most significant impact on product quality. Machine learning models can identify patterns in temperature, nutrient levels, or oxygen conditions to predict how cells will grow. When the system detects a potential problem, it can alert the operator, who can adjust parameters before quality is affected.

Digital twins are virtual models of manufacturing processes that extend these capabilities. They use real-time and historical data to model how changes in variables affect outcomes, allowing scientists to test ideas without interrupting production. Insights gained can lead to fewer failed batches, increased yields, and better utilization of patient-derived materials.

Improve safety and treatment consistency

AI-driven predictive models are improving consistency in manufacturing and patient outcomes and safety. With autologous therapies, each treatment starts with cells from a single patient, and no two samples perform the same. Predictive algorithms can evaluate cellular characteristics to predict how each sample will expand or differentiate. Manufacturers can then adjust culture conditions to maintain potency and viability within target ranges.

In gene therapy, artificial intelligence models are helping to design safer and more predictable viral vectors. These tools can predict gene expression and immune responses, allowing scientists to select ingredients that reduce adverse side effects, improve clinical design, and reduce the risk of late-stage failure.

Predictive control on the manufacturing floor

Manufacturing cell and gene therapies remains one of the most complex undertakings in modern biopharmaceuticals. Each batch can take weeks to complete and can cost hundreds of thousands of dollars. Traditional quality testing typically occurs at the end of the process, which limits the ability to resolve issues that arise early in production.

Advanced analytics and artificial intelligence can now monitor quality in real time. Forecasting systems use data from multiple sources to compare current performance to established models. These systems allow operators to correct problems before they cause failure. This approach supports FDA's Quality by Design principles by embedding quality controls throughout the entire process rather than relying solely on final stage testing.

Predictive control can also improve operational efficiency. By analyzing data from multiple runs, analytical tools can identify which parameters have the greatest impact on yield and turnaround time. Over successive production cycles, this knowledge leads to continuous improvements in cost and reliability.

Challenges Slowing Adoption

Despite the clear potential of analytics and artificial intelligence, its adoption in CGT manufacturing has been incremental. Key barriers include fragmented data systems, limited infrastructure, workforce skills gaps and regulatory uncertainty.

Data fragmentation remains a significant obstacle. Process data, quality indicators, and clinical outcomes often reside in separate databases, preventing a unified view of performance. Without common data standards, even well-designed models cannot easily compare results from different facilities or products.

This challenge is compounded by the lack of a standardized “language” for capital gains tax. Each manufacturer defines process steps, data elements, and parameters differently. Even basic terms such as “viability” or “yield” vary depending on the test or measurement method used. Without a shared vocabulary and data model, it is nearly impossible to align or aggregate data across organizations. Developing this common language is critical to promoting interoperability and enabling meaningful data sharing. Without it, the industry cannot build data sets large enough to effectively train AI systems. Small, siled data sets limit the accuracy and reliability of predictive models, slowing progress toward wider adoption of analytics-based decision-making.

Outdated technology can also slow progress. Many manufacturing plants still rely on instrumentation that lacks connectivity or generates incomplete data sets. Online measurement of critical quality attributes such as cell phenotype or vector potency has not yet been widely utilized. Delayed or missing data can reduce the effectiveness of predictive models. Equipment modernization and digital system upgrades require upfront investment but are critical for long-term scalability.

Process diversity is another challenge. Each CGT product uses different materials and work processes, which limits standardization. A model trained on one platform may not work on another. Developing and maintaining these tools is made more difficult by the lack of professionals who understand bioprocessing and data science.

Regulatory uncertainty continues to impact adoption decisions as companies weigh innovation against compliance risks. Agencies such as the FDA and EMA support innovation in advanced manufacturing but require clear evidence that AI-based systems do not compromise safety or effectiveness. Changes to a validated process may trigger new qualification steps or extend review times. As regulatory frameworks mature, companies will become more confident integrating advanced analytics into production.

Building a digital maturity framework

Overcoming these challenges requires coordination across the CGT ecosystem. Manufacturers, technology providers and regulators can work together to define shared data standards and secure methods for exchanging information. Collaborative initiatives focused on pre-competitive data sharing can provide the large data sets needed to refine predictive models and improve benchmarking.

Infrastructure investment will also accelerate progress. Cloud-based data environments, automated data collection, and integrated manufacturing execution systems make it easier to analyze and act on information. Consistent, high-quality data is the foundation for reliable analytics.

Developing the right workforce is equally important. The field requires professionals who understand cell therapy biology and the computational tools that support cell therapy. Partnerships with universities and training programs can help close this gap and prepare teams for the digital age of manufacturing.

Looking to the future

Advanced analytics and artificial intelligence will not replace human expertise in CGT manufacturing. Instead, they are enhancing it. These tools enable scientists and engineers to make faster, more informed decisions and maintain tighter control over complex processes. Predictive modeling and ongoing monitoring reduce risk, increase efficiency, and help ensure that each patient receives treatment that meets the highest standards of quality and safety.

Digital transformation will be critical as the industry moves from small-scale, patient-specific manufacturing to wider commercial supply. Early adopters of analytics gain a stronger position to maintain consistency and respond to growing market demands. The combination of biology and data science is shaping new standards for advanced treatments and bringing the promise of therapeutic medicine closer to everyday clinical reality.

Photo: Lin Weiquan, Getty Images

This article appeared in Medical City Influencers program. Anyone can share their thoughts on healthcare business and innovation on MedCity News through MedCity Influencers. Click here to learn how.